netdev.c 180.5 KB
Newer Older
1 2 3
/*******************************************************************************

  Intel PRO/1000 Linux driver
B
Bruce Allan 已提交
4
  Copyright(c) 1999 - 2011 Intel Corporation.
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

  This program is free software; you can redistribute it and/or modify it
  under the terms and conditions of the GNU General Public License,
  version 2, as published by the Free Software Foundation.

  This program is distributed in the hope it will be useful, but WITHOUT
  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
  FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
  more details.

  You should have received a copy of the GNU General Public License along with
  this program; if not, write to the Free Software Foundation, Inc.,
  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.

  The full GNU General Public License is included in this distribution in
  the file called "COPYING".

  Contact Information:
  Linux NICS <linux.nics@intel.com>
  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497

*******************************************************************************/

29 30
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

31 32 33 34 35 36 37 38
#include <linux/module.h>
#include <linux/types.h>
#include <linux/init.h>
#include <linux/pci.h>
#include <linux/vmalloc.h>
#include <linux/pagemap.h>
#include <linux/delay.h>
#include <linux/netdevice.h>
39
#include <linux/interrupt.h>
40 41
#include <linux/tcp.h>
#include <linux/ipv6.h>
42
#include <linux/slab.h>
43 44 45 46 47 48 49
#include <net/checksum.h>
#include <net/ip6_checksum.h>
#include <linux/mii.h>
#include <linux/ethtool.h>
#include <linux/if_vlan.h>
#include <linux/cpu.h>
#include <linux/smp.h>
50
#include <linux/pm_qos.h>
51
#include <linux/pm_runtime.h>
J
Jesse Brandeburg 已提交
52
#include <linux/aer.h>
53
#include <linux/prefetch.h>
54 55 56

#include "e1000.h"

B
Bruce Allan 已提交
57
#define DRV_EXTRAVERSION "-k"
58

B
Bruce Allan 已提交
59
#define DRV_VERSION "1.5.1" DRV_EXTRAVERSION
60 61 62
char e1000e_driver_name[] = "e1000e";
const char e1000e_driver_version[] = DRV_VERSION;

63 64
static void e1000e_disable_aspm(struct pci_dev *pdev, u16 state);

65 66 67 68
static const struct e1000_info *e1000_info_tbl[] = {
	[board_82571]		= &e1000_82571_info,
	[board_82572]		= &e1000_82572_info,
	[board_82573]		= &e1000_82573_info,
69
	[board_82574]		= &e1000_82574_info,
70
	[board_82583]		= &e1000_82583_info,
71 72 73
	[board_80003es2lan]	= &e1000_es2_info,
	[board_ich8lan]		= &e1000_ich8_info,
	[board_ich9lan]		= &e1000_ich9_info,
74
	[board_ich10lan]	= &e1000_ich10_info,
75
	[board_pchlan]		= &e1000_pch_info,
76
	[board_pch2lan]		= &e1000_pch2_info,
77 78
};

79 80 81 82 83
struct e1000_reg_info {
	u32 ofs;
	char *name;
};

84 85 86 87 88 89 90 91 92 93 94
#define E1000_RDFH	0x02410	/* Rx Data FIFO Head - RW */
#define E1000_RDFT	0x02418	/* Rx Data FIFO Tail - RW */
#define E1000_RDFHS	0x02420	/* Rx Data FIFO Head Saved - RW */
#define E1000_RDFTS	0x02428	/* Rx Data FIFO Tail Saved - RW */
#define E1000_RDFPC	0x02430	/* Rx Data FIFO Packet Count - RW */

#define E1000_TDFH	0x03410	/* Tx Data FIFO Head - RW */
#define E1000_TDFT	0x03418	/* Tx Data FIFO Tail - RW */
#define E1000_TDFHS	0x03420	/* Tx Data FIFO Head Saved - RW */
#define E1000_TDFTS	0x03428	/* Tx Data FIFO Tail Saved - RW */
#define E1000_TDFPC	0x03430	/* Tx Data FIFO Packet Count - RW */
95 96 97 98 99 100 101 102 103 104 105

static const struct e1000_reg_info e1000_reg_info_tbl[] = {

	/* General Registers */
	{E1000_CTRL, "CTRL"},
	{E1000_STATUS, "STATUS"},
	{E1000_CTRL_EXT, "CTRL_EXT"},

	/* Interrupt Registers */
	{E1000_ICR, "ICR"},

106
	/* Rx Registers */
107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
	{E1000_RCTL, "RCTL"},
	{E1000_RDLEN, "RDLEN"},
	{E1000_RDH, "RDH"},
	{E1000_RDT, "RDT"},
	{E1000_RDTR, "RDTR"},
	{E1000_RXDCTL(0), "RXDCTL"},
	{E1000_ERT, "ERT"},
	{E1000_RDBAL, "RDBAL"},
	{E1000_RDBAH, "RDBAH"},
	{E1000_RDFH, "RDFH"},
	{E1000_RDFT, "RDFT"},
	{E1000_RDFHS, "RDFHS"},
	{E1000_RDFTS, "RDFTS"},
	{E1000_RDFPC, "RDFPC"},

122
	/* Tx Registers */
123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165
	{E1000_TCTL, "TCTL"},
	{E1000_TDBAL, "TDBAL"},
	{E1000_TDBAH, "TDBAH"},
	{E1000_TDLEN, "TDLEN"},
	{E1000_TDH, "TDH"},
	{E1000_TDT, "TDT"},
	{E1000_TIDV, "TIDV"},
	{E1000_TXDCTL(0), "TXDCTL"},
	{E1000_TADV, "TADV"},
	{E1000_TARC(0), "TARC"},
	{E1000_TDFH, "TDFH"},
	{E1000_TDFT, "TDFT"},
	{E1000_TDFHS, "TDFHS"},
	{E1000_TDFTS, "TDFTS"},
	{E1000_TDFPC, "TDFPC"},

	/* List Terminator */
	{}
};

/*
 * e1000_regdump - register printout routine
 */
static void e1000_regdump(struct e1000_hw *hw, struct e1000_reg_info *reginfo)
{
	int n = 0;
	char rname[16];
	u32 regs[8];

	switch (reginfo->ofs) {
	case E1000_RXDCTL(0):
		for (n = 0; n < 2; n++)
			regs[n] = __er32(hw, E1000_RXDCTL(n));
		break;
	case E1000_TXDCTL(0):
		for (n = 0; n < 2; n++)
			regs[n] = __er32(hw, E1000_TXDCTL(n));
		break;
	case E1000_TARC(0):
		for (n = 0; n < 2; n++)
			regs[n] = __er32(hw, E1000_TARC(n));
		break;
	default:
166 167
		pr_info("%-15s %08x\n",
			reginfo->name, __er32(hw, reginfo->ofs));
168 169 170 171
		return;
	}

	snprintf(rname, 16, "%s%s", reginfo->name, "[0-1]");
172
	pr_info("%-15s %08x %08x\n", rname, regs[0], regs[1]);
173 174 175
}

/*
176
 * e1000e_dump - Print registers, Tx-ring and Rx-ring
177 178 179 180 181 182 183 184
 */
static void e1000e_dump(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_reg_info *reginfo;
	struct e1000_ring *tx_ring = adapter->tx_ring;
	struct e1000_tx_desc *tx_desc;
185 186 187 188
	struct my_u0 {
		u64 a;
		u64 b;
	} *u0;
189 190 191
	struct e1000_buffer *buffer_info;
	struct e1000_ring *rx_ring = adapter->rx_ring;
	union e1000_rx_desc_packet_split *rx_desc_ps;
192
	union e1000_rx_desc_extended *rx_desc;
193 194 195 196 197 198
	struct my_u1 {
		u64 a;
		u64 b;
		u64 c;
		u64 d;
	} *u1;
199 200 201 202 203 204 205 206 207
	u32 staterr;
	int i = 0;

	if (!netif_msg_hw(adapter))
		return;

	/* Print netdevice Info */
	if (netdev) {
		dev_info(&adapter->pdev->dev, "Net device Info\n");
208 209 210 211
		pr_info("Device Name     state            trans_start      last_rx\n");
		pr_info("%-15s %016lX %016lX %016lX\n",
			netdev->name, netdev->state, netdev->trans_start,
			netdev->last_rx);
212 213 214 215
	}

	/* Print Registers */
	dev_info(&adapter->pdev->dev, "Register Dump\n");
216
	pr_info(" Register Name   Value\n");
217 218 219 220 221
	for (reginfo = (struct e1000_reg_info *)e1000_reg_info_tbl;
	     reginfo->name; reginfo++) {
		e1000_regdump(hw, reginfo);
	}

222
	/* Print Tx Ring Summary */
223 224 225
	if (!netdev || !netif_running(netdev))
		goto exit;

226
	dev_info(&adapter->pdev->dev, "Tx Ring Summary\n");
227
	pr_info("Queue [NTU] [NTC] [bi(ntc)->dma  ] leng ntw timestamp\n");
228
	buffer_info = &tx_ring->buffer_info[tx_ring->next_to_clean];
229 230 231 232 233 234
	pr_info(" %5d %5X %5X %016llX %04X %3X %016llX\n",
		0, tx_ring->next_to_use, tx_ring->next_to_clean,
		(unsigned long long)buffer_info->dma,
		buffer_info->length,
		buffer_info->next_to_watch,
		(unsigned long long)buffer_info->time_stamp);
235

236
	/* Print Tx Ring */
237 238 239
	if (!netif_msg_tx_done(adapter))
		goto rx_ring_summary;

240
	dev_info(&adapter->pdev->dev, "Tx Ring Dump\n");
241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268

	/* Transmit Descriptor Formats - DEXT[29] is 0 (Legacy) or 1 (Extended)
	 *
	 * Legacy Transmit Descriptor
	 *   +--------------------------------------------------------------+
	 * 0 |         Buffer Address [63:0] (Reserved on Write Back)       |
	 *   +--------------------------------------------------------------+
	 * 8 | Special  |    CSS     | Status |  CMD    |  CSO   |  Length  |
	 *   +--------------------------------------------------------------+
	 *   63       48 47        36 35    32 31     24 23    16 15        0
	 *
	 * Extended Context Descriptor (DTYP=0x0) for TSO or checksum offload
	 *   63      48 47    40 39       32 31             16 15    8 7      0
	 *   +----------------------------------------------------------------+
	 * 0 |  TUCSE  | TUCS0  |   TUCSS   |     IPCSE       | IPCS0 | IPCSS |
	 *   +----------------------------------------------------------------+
	 * 8 |   MSS   | HDRLEN | RSV | STA | TUCMD | DTYP |      PAYLEN      |
	 *   +----------------------------------------------------------------+
	 *   63      48 47    40 39 36 35 32 31   24 23  20 19                0
	 *
	 * Extended Data Descriptor (DTYP=0x1)
	 *   +----------------------------------------------------------------+
	 * 0 |                     Buffer Address [63:0]                      |
	 *   +----------------------------------------------------------------+
	 * 8 | VLAN tag |  POPTS  | Rsvd | Status | Command | DTYP |  DTALEN  |
	 *   +----------------------------------------------------------------+
	 *   63       48 47     40 39  36 35    32 31     24 23  20 19        0
	 */
269 270 271
	pr_info("Tl[desc]     [address 63:0  ] [SpeCssSCmCsLen] [bi->dma       ] leng  ntw timestamp        bi->skb <-- Legacy format\n");
	pr_info("Tc[desc]     [Ce CoCsIpceCoS] [MssHlRSCm0Plen] [bi->dma       ] leng  ntw timestamp        bi->skb <-- Ext Context format\n");
	pr_info("Td[desc]     [address 63:0  ] [VlaPoRSCm1Dlen] [bi->dma       ] leng  ntw timestamp        bi->skb <-- Ext Data format\n");
272
	for (i = 0; tx_ring->desc && (i < tx_ring->count); i++) {
273
		const char *next_desc;
274 275 276 277
		tx_desc = E1000_TX_DESC(*tx_ring, i);
		buffer_info = &tx_ring->buffer_info[i];
		u0 = (struct my_u0 *)tx_desc;
		if (i == tx_ring->next_to_use && i == tx_ring->next_to_clean)
278
			next_desc = " NTC/U";
279
		else if (i == tx_ring->next_to_use)
280
			next_desc = " NTU";
281
		else if (i == tx_ring->next_to_clean)
282
			next_desc = " NTC";
283
		else
284 285 286 287 288 289 290 291 292 293 294
			next_desc = "";
		pr_info("T%c[0x%03X]    %016llX %016llX %016llX %04X  %3X %016llX %p%s\n",
			(!(le64_to_cpu(u0->b) & (1 << 29)) ? 'l' :
			 ((le64_to_cpu(u0->b) & (1 << 20)) ? 'd' : 'c')),
			i,
			(unsigned long long)le64_to_cpu(u0->a),
			(unsigned long long)le64_to_cpu(u0->b),
			(unsigned long long)buffer_info->dma,
			buffer_info->length, buffer_info->next_to_watch,
			(unsigned long long)buffer_info->time_stamp,
			buffer_info->skb, next_desc);
295 296 297

		if (netif_msg_pktdata(adapter) && buffer_info->dma != 0)
			print_hex_dump(KERN_INFO, "", DUMP_PREFIX_ADDRESS,
298 299
				       16, 1, phys_to_virt(buffer_info->dma),
				       buffer_info->length, true);
300 301
	}

302
	/* Print Rx Ring Summary */
303
rx_ring_summary:
304
	dev_info(&adapter->pdev->dev, "Rx Ring Summary\n");
305 306 307
	pr_info("Queue [NTU] [NTC]\n");
	pr_info(" %5d %5X %5X\n",
		0, rx_ring->next_to_use, rx_ring->next_to_clean);
308

309
	/* Print Rx Ring */
310 311 312
	if (!netif_msg_rx_status(adapter))
		goto exit;

313
	dev_info(&adapter->pdev->dev, "Rx Ring Dump\n");
314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329
	switch (adapter->rx_ps_pages) {
	case 1:
	case 2:
	case 3:
		/* [Extended] Packet Split Receive Descriptor Format
		 *
		 *    +-----------------------------------------------------+
		 *  0 |                Buffer Address 0 [63:0]              |
		 *    +-----------------------------------------------------+
		 *  8 |                Buffer Address 1 [63:0]              |
		 *    +-----------------------------------------------------+
		 * 16 |                Buffer Address 2 [63:0]              |
		 *    +-----------------------------------------------------+
		 * 24 |                Buffer Address 3 [63:0]              |
		 *    +-----------------------------------------------------+
		 */
330
		pr_info("R  [desc]      [buffer 0 63:0 ] [buffer 1 63:0 ] [buffer 2 63:0 ] [buffer 3 63:0 ] [bi->dma       ] [bi->skb] <-- Ext Pkt Split format\n");
331 332 333 334 335 336 337 338 339 340 341
		/* [Extended] Receive Descriptor (Write-Back) Format
		 *
		 *   63       48 47    32 31     13 12    8 7    4 3        0
		 *   +------------------------------------------------------+
		 * 0 | Packet   | IP     |  Rsvd   | MRQ   | Rsvd | MRQ RSS |
		 *   | Checksum | Ident  |         | Queue |      |  Type   |
		 *   +------------------------------------------------------+
		 * 8 | VLAN Tag | Length | Extended Error | Extended Status |
		 *   +------------------------------------------------------+
		 *   63       48 47    32 31            20 19               0
		 */
342
		pr_info("RWB[desc]      [ck ipid mrqhsh] [vl   l0 ee  es] [ l3  l2  l1 hs] [reserved      ] ---------------- [bi->skb] <-- Ext Rx Write-Back format\n");
343
		for (i = 0; i < rx_ring->count; i++) {
344
			const char *next_desc;
345 346 347 348
			buffer_info = &rx_ring->buffer_info[i];
			rx_desc_ps = E1000_RX_DESC_PS(*rx_ring, i);
			u1 = (struct my_u1 *)rx_desc_ps;
			staterr =
349
			    le32_to_cpu(rx_desc_ps->wb.middle.status_error);
350 351 352 353 354 355 356 357

			if (i == rx_ring->next_to_use)
				next_desc = " NTU";
			else if (i == rx_ring->next_to_clean)
				next_desc = " NTC";
			else
				next_desc = "";

358 359
			if (staterr & E1000_RXD_STAT_DD) {
				/* Descriptor Done */
360 361 362 363 364 365 366
				pr_info("%s[0x%03X]     %016llX %016llX %016llX %016llX ---------------- %p%s\n",
					"RWB", i,
					(unsigned long long)le64_to_cpu(u1->a),
					(unsigned long long)le64_to_cpu(u1->b),
					(unsigned long long)le64_to_cpu(u1->c),
					(unsigned long long)le64_to_cpu(u1->d),
					buffer_info->skb, next_desc);
367
			} else {
368 369 370 371 372 373 374 375
				pr_info("%s[0x%03X]     %016llX %016llX %016llX %016llX %016llX %p%s\n",
					"R  ", i,
					(unsigned long long)le64_to_cpu(u1->a),
					(unsigned long long)le64_to_cpu(u1->b),
					(unsigned long long)le64_to_cpu(u1->c),
					(unsigned long long)le64_to_cpu(u1->d),
					(unsigned long long)buffer_info->dma,
					buffer_info->skb, next_desc);
376 377 378 379 380 381 382 383 384 385 386

				if (netif_msg_pktdata(adapter))
					print_hex_dump(KERN_INFO, "",
						DUMP_PREFIX_ADDRESS, 16, 1,
						phys_to_virt(buffer_info->dma),
						adapter->rx_ps_bsize0, true);
			}
		}
		break;
	default:
	case 0:
387
		/* Extended Receive Descriptor (Read) Format
388
		 *
389 390 391 392 393
		 *   +-----------------------------------------------------+
		 * 0 |                Buffer Address [63:0]                |
		 *   +-----------------------------------------------------+
		 * 8 |                      Reserved                       |
		 *   +-----------------------------------------------------+
394
		 */
395
		pr_info("R  [desc]      [buf addr 63:0 ] [reserved 63:0 ] [bi->dma       ] [bi->skb] <-- Ext (Read) format\n");
396 397 398 399 400 401 402 403 404 405 406 407 408
		/* Extended Receive Descriptor (Write-Back) Format
		 *
		 *   63       48 47    32 31    24 23            4 3        0
		 *   +------------------------------------------------------+
		 *   |     RSS Hash      |        |               |         |
		 * 0 +-------------------+  Rsvd  |   Reserved    | MRQ RSS |
		 *   | Packet   | IP     |        |               |  Type   |
		 *   | Checksum | Ident  |        |               |         |
		 *   +------------------------------------------------------+
		 * 8 | VLAN Tag | Length | Extended Error | Extended Status |
		 *   +------------------------------------------------------+
		 *   63       48 47    32 31            20 19               0
		 */
409
		pr_info("RWB[desc]      [cs ipid    mrq] [vt   ln xe  xs] [bi->skb] <-- Ext (Write-Back) format\n");
410 411

		for (i = 0; i < rx_ring->count; i++) {
412 413
			const char *next_desc;

414
			buffer_info = &rx_ring->buffer_info[i];
415 416 417
			rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
			u1 = (struct my_u1 *)rx_desc;
			staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
418 419 420 421 422 423 424 425

			if (i == rx_ring->next_to_use)
				next_desc = " NTU";
			else if (i == rx_ring->next_to_clean)
				next_desc = " NTC";
			else
				next_desc = "";

426 427
			if (staterr & E1000_RXD_STAT_DD) {
				/* Descriptor Done */
428 429 430 431 432
				pr_info("%s[0x%03X]     %016llX %016llX ---------------- %p%s\n",
					"RWB", i,
					(unsigned long long)le64_to_cpu(u1->a),
					(unsigned long long)le64_to_cpu(u1->b),
					buffer_info->skb, next_desc);
433
			} else {
434 435 436 437 438 439
				pr_info("%s[0x%03X]     %016llX %016llX %016llX %p%s\n",
					"R  ", i,
					(unsigned long long)le64_to_cpu(u1->a),
					(unsigned long long)le64_to_cpu(u1->b),
					(unsigned long long)buffer_info->dma,
					buffer_info->skb, next_desc);
440 441 442 443 444 445 446 447 448 449

				if (netif_msg_pktdata(adapter))
					print_hex_dump(KERN_INFO, "",
						       DUMP_PREFIX_ADDRESS, 16,
						       1,
						       phys_to_virt
						       (buffer_info->dma),
						       adapter->rx_buffer_len,
						       true);
			}
450 451 452 453 454 455 456
		}
	}

exit:
	return;
}

457 458 459 460 461 462 463 464 465 466 467 468
/**
 * e1000_desc_unused - calculate if we have unused descriptors
 **/
static int e1000_desc_unused(struct e1000_ring *ring)
{
	if (ring->next_to_clean > ring->next_to_use)
		return ring->next_to_clean - ring->next_to_use - 1;

	return ring->count + ring->next_to_clean - ring->next_to_use - 1;
}

/**
469
 * e1000_receive_skb - helper function to handle Rx indications
470 471 472 473 474 475
 * @adapter: board private structure
 * @status: descriptor status field as written by hardware
 * @vlan: descriptor vlan field as written by hardware (no le/be conversion)
 * @skb: pointer to sk_buff to be indicated to stack
 **/
static void e1000_receive_skb(struct e1000_adapter *adapter,
476
			      struct net_device *netdev, struct sk_buff *skb,
A
Al Viro 已提交
477
			      u8 status, __le16 vlan)
478
{
J
Jeff Kirsher 已提交
479
	u16 tag = le16_to_cpu(vlan);
480 481
	skb->protocol = eth_type_trans(skb, netdev);

J
Jeff Kirsher 已提交
482 483 484 485
	if (status & E1000_RXD_STAT_VP)
		__vlan_hwaccel_put_tag(skb, tag);

	napi_gro_receive(&adapter->napi, skb);
486 487 488
}

/**
489
 * e1000_rx_checksum - Receive Checksum Offload
490 491 492 493
 * @adapter: board private structure
 * @status_err: receive descriptor status and error fields
 * @csum: receive descriptor csum field
 * @sk_buff: socket buffer with received data
494 495
 **/
static void e1000_rx_checksum(struct e1000_adapter *adapter, u32 status_err,
496
			      __le16 csum, struct sk_buff *skb)
497 498 499
{
	u16 status = (u16)status_err;
	u8 errors = (u8)(status_err >> 24);
500 501

	skb_checksum_none_assert(skb);
502

503 504 505 506
	/* Rx checksum disabled */
	if (!(adapter->netdev->features & NETIF_F_RXCSUM))
		return;

507 508 509
	/* Ignore Checksum bit is set */
	if (status & E1000_RXD_STAT_IXSM)
		return;
510

511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526
	/* TCP/UDP checksum error bit is set */
	if (errors & E1000_RXD_ERR_TCPE) {
		/* let the stack verify checksum errors */
		adapter->hw_csum_err++;
		return;
	}

	/* TCP/UDP Checksum has not been calculated */
	if (!(status & (E1000_RXD_STAT_TCPCS | E1000_RXD_STAT_UDPCS)))
		return;

	/* It must be a TCP or UDP packet with a valid checksum */
	if (status & E1000_RXD_STAT_TCPCS) {
		/* TCP checksum is good */
		skb->ip_summed = CHECKSUM_UNNECESSARY;
	} else {
527 528 529
		/*
		 * IP fragment with UDP payload
		 * Hardware complements the payload checksum, so we undo it
530 531
		 * and then put the value in host order for further stack use.
		 */
532
		__sum16 sum = (__force __sum16)swab16((__force u16)csum);
A
Al Viro 已提交
533
		skb->csum = csum_unfold(~sum);
534 535 536 537 538
		skb->ip_summed = CHECKSUM_COMPLETE;
	}
	adapter->hw_csum_good++;
}

539 540 541 542 543 544 545 546 547 548 549 550 551 552
/**
 * e1000e_update_tail_wa - helper function for e1000e_update_[rt]dt_wa()
 * @hw: pointer to the HW structure
 * @tail: address of tail descriptor register
 * @i: value to write to tail descriptor register
 *
 * When updating the tail register, the ME could be accessing Host CSR
 * registers at the same time.  Normally, this is handled in h/w by an
 * arbiter but on some parts there is a bug that acknowledges Host accesses
 * later than it should which could result in the descriptor register to
 * have an incorrect value.  Workaround this by checking the FWSM register
 * which has bit 24 set while ME is accessing Host CSR registers, wait
 * if it is set and try again a number of times.
 **/
553
static inline s32 e1000e_update_tail_wa(struct e1000_hw *hw, void __iomem *tail,
554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569
					unsigned int i)
{
	unsigned int j = 0;

	while ((j++ < E1000_ICH_FWSM_PCIM2PCI_COUNT) &&
	       (er32(FWSM) & E1000_ICH_FWSM_PCIM2PCI))
		udelay(50);

	writel(i, tail);

	if ((j == E1000_ICH_FWSM_PCIM2PCI_COUNT) && (i != readl(tail)))
		return E1000_ERR_SWFW_SYNC;

	return 0;
}

570
static void e1000e_update_rdt_wa(struct e1000_ring *rx_ring, unsigned int i)
571
{
572
	struct e1000_adapter *adapter = rx_ring->adapter;
573 574
	struct e1000_hw *hw = &adapter->hw;

575
	if (e1000e_update_tail_wa(hw, rx_ring->tail, i)) {
576 577 578 579 580 581 582
		u32 rctl = er32(RCTL);
		ew32(RCTL, rctl & ~E1000_RCTL_EN);
		e_err("ME firmware caused invalid RDT - resetting\n");
		schedule_work(&adapter->reset_task);
	}
}

583
static void e1000e_update_tdt_wa(struct e1000_ring *tx_ring, unsigned int i)
584
{
585
	struct e1000_adapter *adapter = tx_ring->adapter;
586 587
	struct e1000_hw *hw = &adapter->hw;

588
	if (e1000e_update_tail_wa(hw, tx_ring->tail, i)) {
589 590 591 592 593 594 595
		u32 tctl = er32(TCTL);
		ew32(TCTL, tctl & ~E1000_TCTL_EN);
		e_err("ME firmware caused invalid TDT - resetting\n");
		schedule_work(&adapter->reset_task);
	}
}

596
/**
597
 * e1000_alloc_rx_buffers - Replace used receive buffers
598
 * @rx_ring: Rx descriptor ring
599
 **/
600
static void e1000_alloc_rx_buffers(struct e1000_ring *rx_ring,
601
				   int cleaned_count, gfp_t gfp)
602
{
603
	struct e1000_adapter *adapter = rx_ring->adapter;
604 605
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
606
	union e1000_rx_desc_extended *rx_desc;
607 608 609
	struct e1000_buffer *buffer_info;
	struct sk_buff *skb;
	unsigned int i;
610
	unsigned int bufsz = adapter->rx_buffer_len;
611 612 613 614 615 616 617 618 619 620 621

	i = rx_ring->next_to_use;
	buffer_info = &rx_ring->buffer_info[i];

	while (cleaned_count--) {
		skb = buffer_info->skb;
		if (skb) {
			skb_trim(skb, 0);
			goto map_skb;
		}

622
		skb = __netdev_alloc_skb_ip_align(netdev, bufsz, gfp);
623 624 625 626 627 628 629 630
		if (!skb) {
			/* Better luck next round */
			adapter->alloc_rx_buff_failed++;
			break;
		}

		buffer_info->skb = skb;
map_skb:
631
		buffer_info->dma = dma_map_single(&pdev->dev, skb->data,
632
						  adapter->rx_buffer_len,
633 634
						  DMA_FROM_DEVICE);
		if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
635
			dev_err(&pdev->dev, "Rx DMA map failed\n");
636 637 638 639
			adapter->rx_dma_failed++;
			break;
		}

640 641
		rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
		rx_desc->read.buffer_addr = cpu_to_le64(buffer_info->dma);
642

643 644 645 646 647 648 649 650
		if (unlikely(!(i & (E1000_RX_BUFFER_WRITE - 1)))) {
			/*
			 * Force memory writes to complete before letting h/w
			 * know there are new descriptors to fetch.  (Only
			 * applicable for weak-ordered memory model archs,
			 * such as IA-64).
			 */
			wmb();
651
			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
652
				e1000e_update_rdt_wa(rx_ring, i);
653
			else
654
				writel(i, rx_ring->tail);
655
		}
656 657 658 659 660 661
		i++;
		if (i == rx_ring->count)
			i = 0;
		buffer_info = &rx_ring->buffer_info[i];
	}

662
	rx_ring->next_to_use = i;
663 664 665 666
}

/**
 * e1000_alloc_rx_buffers_ps - Replace used receive buffers; packet split
667
 * @rx_ring: Rx descriptor ring
668
 **/
669
static void e1000_alloc_rx_buffers_ps(struct e1000_ring *rx_ring,
670
				      int cleaned_count, gfp_t gfp)
671
{
672
	struct e1000_adapter *adapter = rx_ring->adapter;
673 674 675 676 677 678 679 680 681 682 683 684 685 686 687
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
	union e1000_rx_desc_packet_split *rx_desc;
	struct e1000_buffer *buffer_info;
	struct e1000_ps_page *ps_page;
	struct sk_buff *skb;
	unsigned int i, j;

	i = rx_ring->next_to_use;
	buffer_info = &rx_ring->buffer_info[i];

	while (cleaned_count--) {
		rx_desc = E1000_RX_DESC_PS(*rx_ring, i);

		for (j = 0; j < PS_PAGE_BUFFERS; j++) {
A
Auke Kok 已提交
688 689 690
			ps_page = &buffer_info->ps_pages[j];
			if (j >= adapter->rx_ps_pages) {
				/* all unused desc entries get hw null ptr */
691 692
				rx_desc->read.buffer_addr[j + 1] =
				    ~cpu_to_le64(0);
A
Auke Kok 已提交
693 694 695
				continue;
			}
			if (!ps_page->page) {
696
				ps_page->page = alloc_page(gfp);
697
				if (!ps_page->page) {
A
Auke Kok 已提交
698 699 700
					adapter->alloc_rx_buff_failed++;
					goto no_buffers;
				}
701 702 703 704 705 706
				ps_page->dma = dma_map_page(&pdev->dev,
							    ps_page->page,
							    0, PAGE_SIZE,
							    DMA_FROM_DEVICE);
				if (dma_mapping_error(&pdev->dev,
						      ps_page->dma)) {
A
Auke Kok 已提交
707
					dev_err(&adapter->pdev->dev,
708
						"Rx DMA page map failed\n");
A
Auke Kok 已提交
709 710
					adapter->rx_dma_failed++;
					goto no_buffers;
711 712
				}
			}
A
Auke Kok 已提交
713 714 715 716 717
			/*
			 * Refresh the desc even if buffer_addrs
			 * didn't change because each write-back
			 * erases this info.
			 */
718 719
			rx_desc->read.buffer_addr[j + 1] =
			    cpu_to_le64(ps_page->dma);
720 721
		}

722 723 724
		skb = __netdev_alloc_skb_ip_align(netdev,
						  adapter->rx_ps_bsize0,
						  gfp);
725 726 727 728 729 730 731

		if (!skb) {
			adapter->alloc_rx_buff_failed++;
			break;
		}

		buffer_info->skb = skb;
732
		buffer_info->dma = dma_map_single(&pdev->dev, skb->data,
733
						  adapter->rx_ps_bsize0,
734 735
						  DMA_FROM_DEVICE);
		if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
736
			dev_err(&pdev->dev, "Rx DMA map failed\n");
737 738 739 740 741 742 743 744 745
			adapter->rx_dma_failed++;
			/* cleanup skb */
			dev_kfree_skb_any(skb);
			buffer_info->skb = NULL;
			break;
		}

		rx_desc->read.buffer_addr[0] = cpu_to_le64(buffer_info->dma);

746 747 748 749 750 751 752 753
		if (unlikely(!(i & (E1000_RX_BUFFER_WRITE - 1)))) {
			/*
			 * Force memory writes to complete before letting h/w
			 * know there are new descriptors to fetch.  (Only
			 * applicable for weak-ordered memory model archs,
			 * such as IA-64).
			 */
			wmb();
754
			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
755
				e1000e_update_rdt_wa(rx_ring, i << 1);
756
			else
757
				writel(i << 1, rx_ring->tail);
758 759
		}

760 761 762 763 764 765 766
		i++;
		if (i == rx_ring->count)
			i = 0;
		buffer_info = &rx_ring->buffer_info[i];
	}

no_buffers:
767
	rx_ring->next_to_use = i;
768 769
}

770 771
/**
 * e1000_alloc_jumbo_rx_buffers - Replace used jumbo receive buffers
772
 * @rx_ring: Rx descriptor ring
773 774 775
 * @cleaned_count: number of buffers to allocate this pass
 **/

776
static void e1000_alloc_jumbo_rx_buffers(struct e1000_ring *rx_ring,
777
					 int cleaned_count, gfp_t gfp)
778
{
779
	struct e1000_adapter *adapter = rx_ring->adapter;
780 781
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
782
	union e1000_rx_desc_extended *rx_desc;
783 784 785
	struct e1000_buffer *buffer_info;
	struct sk_buff *skb;
	unsigned int i;
786
	unsigned int bufsz = 256 - 16 /* for skb_reserve */;
787 788 789 790 791 792 793 794 795 796 797

	i = rx_ring->next_to_use;
	buffer_info = &rx_ring->buffer_info[i];

	while (cleaned_count--) {
		skb = buffer_info->skb;
		if (skb) {
			skb_trim(skb, 0);
			goto check_page;
		}

798
		skb = __netdev_alloc_skb_ip_align(netdev, bufsz, gfp);
799 800 801 802 803 804 805 806 807 808
		if (unlikely(!skb)) {
			/* Better luck next round */
			adapter->alloc_rx_buff_failed++;
			break;
		}

		buffer_info->skb = skb;
check_page:
		/* allocate a new page if necessary */
		if (!buffer_info->page) {
809
			buffer_info->page = alloc_page(gfp);
810 811 812 813 814 815 816
			if (unlikely(!buffer_info->page)) {
				adapter->alloc_rx_buff_failed++;
				break;
			}
		}

		if (!buffer_info->dma)
817
			buffer_info->dma = dma_map_page(&pdev->dev,
818 819
			                                buffer_info->page, 0,
			                                PAGE_SIZE,
820
							DMA_FROM_DEVICE);
821

822 823
		rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
		rx_desc->read.buffer_addr = cpu_to_le64(buffer_info->dma);
824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839

		if (unlikely(++i == rx_ring->count))
			i = 0;
		buffer_info = &rx_ring->buffer_info[i];
	}

	if (likely(rx_ring->next_to_use != i)) {
		rx_ring->next_to_use = i;
		if (unlikely(i-- == 0))
			i = (rx_ring->count - 1);

		/* Force memory writes to complete before letting h/w
		 * know there are new descriptors to fetch.  (Only
		 * applicable for weak-ordered memory model archs,
		 * such as IA-64). */
		wmb();
840
		if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
841
			e1000e_update_rdt_wa(rx_ring, i);
842
		else
843
			writel(i, rx_ring->tail);
844 845 846
	}
}

847 848 849 850 851 852 853
static inline void e1000_rx_hash(struct net_device *netdev, __le32 rss,
				 struct sk_buff *skb)
{
	if (netdev->features & NETIF_F_RXHASH)
		skb->rxhash = le32_to_cpu(rss);
}

854
/**
855 856
 * e1000_clean_rx_irq - Send received data up the network stack
 * @rx_ring: Rx descriptor ring
857 858 859 860
 *
 * the return value indicates whether actual cleaning was done, there
 * is no guarantee that everything was cleaned
 **/
861 862
static bool e1000_clean_rx_irq(struct e1000_ring *rx_ring, int *work_done,
			       int work_to_do)
863
{
864
	struct e1000_adapter *adapter = rx_ring->adapter;
865 866
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
867
	struct e1000_hw *hw = &adapter->hw;
868
	union e1000_rx_desc_extended *rx_desc, *next_rxd;
869
	struct e1000_buffer *buffer_info, *next_buffer;
870
	u32 length, staterr;
871 872
	unsigned int i;
	int cleaned_count = 0;
873
	bool cleaned = false;
874 875 876
	unsigned int total_rx_bytes = 0, total_rx_packets = 0;

	i = rx_ring->next_to_clean;
877 878
	rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
	staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
879 880
	buffer_info = &rx_ring->buffer_info[i];

881
	while (staterr & E1000_RXD_STAT_DD) {
882 883 884 885 886
		struct sk_buff *skb;

		if (*work_done >= work_to_do)
			break;
		(*work_done)++;
887
		rmb();	/* read descriptor and rx_buffer_info after status DD */
888 889 890 891 892 893 894 895 896

		skb = buffer_info->skb;
		buffer_info->skb = NULL;

		prefetch(skb->data - NET_IP_ALIGN);

		i++;
		if (i == rx_ring->count)
			i = 0;
897
		next_rxd = E1000_RX_DESC_EXT(*rx_ring, i);
898 899 900 901
		prefetch(next_rxd);

		next_buffer = &rx_ring->buffer_info[i];

902
		cleaned = true;
903
		cleaned_count++;
904
		dma_unmap_single(&pdev->dev,
905 906
				 buffer_info->dma,
				 adapter->rx_buffer_len,
907
				 DMA_FROM_DEVICE);
908 909
		buffer_info->dma = 0;

910
		length = le16_to_cpu(rx_desc->wb.upper.length);
911

912 913 914 915 916 917 918
		/*
		 * !EOP means multiple descriptors were used to store a single
		 * packet, if that's the case we need to toss it.  In fact, we
		 * need to toss every packet with the EOP bit clear and the
		 * next frame that _does_ have the EOP bit set, as it is by
		 * definition only a frame fragment
		 */
919
		if (unlikely(!(staterr & E1000_RXD_STAT_EOP)))
920 921 922
			adapter->flags2 |= FLAG2_IS_DISCARDING;

		if (adapter->flags2 & FLAG2_IS_DISCARDING) {
923
			/* All receives must fit into a single buffer */
924
			e_dbg("Receive packet consumed multiple buffers\n");
925 926
			/* recycle */
			buffer_info->skb = skb;
927
			if (staterr & E1000_RXD_STAT_EOP)
928
				adapter->flags2 &= ~FLAG2_IS_DISCARDING;
929 930 931
			goto next_desc;
		}

932
		if (staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) {
933 934 935 936 937
			/* recycle */
			buffer_info->skb = skb;
			goto next_desc;
		}

J
Jeff Kirsher 已提交
938 939 940 941
		/* adjust length to remove Ethernet CRC */
		if (!(adapter->flags2 & FLAG2_CRC_STRIPPING))
			length -= 4;

942 943 944
		total_rx_bytes += length;
		total_rx_packets++;

945 946
		/*
		 * code added for copybreak, this should improve
947
		 * performance for small packets with large amounts
948 949
		 * of reassembly being done in the stack
		 */
950 951
		if (length < copybreak) {
			struct sk_buff *new_skb =
952
			    netdev_alloc_skb_ip_align(netdev, length);
953
			if (new_skb) {
954 955 956 957 958 959
				skb_copy_to_linear_data_offset(new_skb,
							       -NET_IP_ALIGN,
							       (skb->data -
								NET_IP_ALIGN),
							       (length +
								NET_IP_ALIGN));
960 961 962 963 964 965 966 967 968 969
				/* save the skb in buffer_info as good */
				buffer_info->skb = skb;
				skb = new_skb;
			}
			/* else just continue with the old one */
		}
		/* end copybreak code */
		skb_put(skb, length);

		/* Receive Checksum Offload */
970
		e1000_rx_checksum(adapter, staterr,
971
				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
972

973 974
		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);

975 976
		e1000_receive_skb(adapter, netdev, skb, staterr,
				  rx_desc->wb.upper.vlan);
977 978

next_desc:
979
		rx_desc->wb.upper.status_error &= cpu_to_le32(~0xFF);
980 981 982

		/* return some buffers to hardware, one at a time is too slow */
		if (cleaned_count >= E1000_RX_BUFFER_WRITE) {
983
			adapter->alloc_rx_buf(rx_ring, cleaned_count,
984
					      GFP_ATOMIC);
985 986 987 988 989 990
			cleaned_count = 0;
		}

		/* use prefetched values */
		rx_desc = next_rxd;
		buffer_info = next_buffer;
991 992

		staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
993 994 995 996 997
	}
	rx_ring->next_to_clean = i;

	cleaned_count = e1000_desc_unused(rx_ring);
	if (cleaned_count)
998
		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
999 1000

	adapter->total_rx_bytes += total_rx_bytes;
1001
	adapter->total_rx_packets += total_rx_packets;
1002 1003 1004
	return cleaned;
}

1005 1006
static void e1000_put_txbuf(struct e1000_ring *tx_ring,
			    struct e1000_buffer *buffer_info)
1007
{
1008 1009
	struct e1000_adapter *adapter = tx_ring->adapter;

1010 1011
	if (buffer_info->dma) {
		if (buffer_info->mapped_as_page)
1012 1013
			dma_unmap_page(&adapter->pdev->dev, buffer_info->dma,
				       buffer_info->length, DMA_TO_DEVICE);
1014
		else
1015 1016
			dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
					 buffer_info->length, DMA_TO_DEVICE);
1017 1018
		buffer_info->dma = 0;
	}
1019 1020 1021 1022
	if (buffer_info->skb) {
		dev_kfree_skb_any(buffer_info->skb);
		buffer_info->skb = NULL;
	}
1023
	buffer_info->time_stamp = 0;
1024 1025
}

1026
static void e1000_print_hw_hang(struct work_struct *work)
1027
{
1028 1029 1030
	struct e1000_adapter *adapter = container_of(work,
	                                             struct e1000_adapter,
	                                             print_hang_task);
1031
	struct net_device *netdev = adapter->netdev;
1032 1033 1034 1035
	struct e1000_ring *tx_ring = adapter->tx_ring;
	unsigned int i = tx_ring->next_to_clean;
	unsigned int eop = tx_ring->buffer_info[i].next_to_watch;
	struct e1000_tx_desc *eop_desc = E1000_TX_DESC(*tx_ring, eop);
1036 1037 1038 1039
	struct e1000_hw *hw = &adapter->hw;
	u16 phy_status, phy_1000t_status, phy_ext_status;
	u16 pci_status;

1040 1041 1042
	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057
	if (!adapter->tx_hang_recheck &&
	    (adapter->flags2 & FLAG2_DMA_BURST)) {
		/* May be block on write-back, flush and detect again
		 * flush pending descriptor writebacks to memory
		 */
		ew32(TIDV, adapter->tx_int_delay | E1000_TIDV_FPD);
		/* execute the writes immediately */
		e1e_flush();
		adapter->tx_hang_recheck = true;
		return;
	}
	/* Real hang detected */
	adapter->tx_hang_recheck = false;
	netif_stop_queue(netdev);

1058 1059 1060
	e1e_rphy(hw, PHY_STATUS, &phy_status);
	e1e_rphy(hw, PHY_1000T_STATUS, &phy_1000t_status);
	e1e_rphy(hw, PHY_EXT_STATUS, &phy_ext_status);
1061

1062 1063 1064 1065
	pci_read_config_word(adapter->pdev, PCI_STATUS, &pci_status);

	/* detected Hardware unit hang */
	e_err("Detected Hardware Unit Hang:\n"
1066 1067 1068 1069 1070 1071 1072 1073
	      "  TDH                  <%x>\n"
	      "  TDT                  <%x>\n"
	      "  next_to_use          <%x>\n"
	      "  next_to_clean        <%x>\n"
	      "buffer_info[next_to_clean]:\n"
	      "  time_stamp           <%lx>\n"
	      "  next_to_watch        <%x>\n"
	      "  jiffies              <%lx>\n"
1074 1075 1076 1077 1078 1079
	      "  next_to_watch.status <%x>\n"
	      "MAC Status             <%x>\n"
	      "PHY Status             <%x>\n"
	      "PHY 1000BASE-T Status  <%x>\n"
	      "PHY Extended Status    <%x>\n"
	      "PCI Status             <%x>\n",
1080 1081
	      readl(tx_ring->head),
	      readl(tx_ring->tail),
1082 1083 1084 1085 1086
	      tx_ring->next_to_use,
	      tx_ring->next_to_clean,
	      tx_ring->buffer_info[eop].time_stamp,
	      eop,
	      jiffies,
1087 1088 1089 1090 1091 1092
	      eop_desc->upper.fields.status,
	      er32(STATUS),
	      phy_status,
	      phy_1000t_status,
	      phy_ext_status,
	      pci_status);
1093 1094 1095 1096
}

/**
 * e1000_clean_tx_irq - Reclaim resources after transmit completes
1097
 * @tx_ring: Tx descriptor ring
1098 1099 1100 1101
 *
 * the return value indicates whether actual cleaning was done, there
 * is no guarantee that everything was cleaned
 **/
1102
static bool e1000_clean_tx_irq(struct e1000_ring *tx_ring)
1103
{
1104
	struct e1000_adapter *adapter = tx_ring->adapter;
1105 1106 1107 1108 1109 1110 1111
	struct net_device *netdev = adapter->netdev;
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_tx_desc *tx_desc, *eop_desc;
	struct e1000_buffer *buffer_info;
	unsigned int i, eop;
	unsigned int count = 0;
	unsigned int total_tx_bytes = 0, total_tx_packets = 0;
1112
	unsigned int bytes_compl = 0, pkts_compl = 0;
1113 1114 1115 1116 1117

	i = tx_ring->next_to_clean;
	eop = tx_ring->buffer_info[i].next_to_watch;
	eop_desc = E1000_TX_DESC(*tx_ring, eop);

1118 1119
	while ((eop_desc->upper.data & cpu_to_le32(E1000_TXD_STAT_DD)) &&
	       (count < tx_ring->count)) {
1120
		bool cleaned = false;
1121
		rmb(); /* read buffer_info after eop_desc */
1122
		for (; !cleaned; count++) {
1123 1124 1125 1126 1127
			tx_desc = E1000_TX_DESC(*tx_ring, i);
			buffer_info = &tx_ring->buffer_info[i];
			cleaned = (i == eop);

			if (cleaned) {
1128 1129
				total_tx_packets += buffer_info->segs;
				total_tx_bytes += buffer_info->bytecount;
1130 1131 1132 1133
				if (buffer_info->skb) {
					bytes_compl += buffer_info->skb->len;
					pkts_compl++;
				}
1134 1135
			}

1136
			e1000_put_txbuf(tx_ring, buffer_info);
1137 1138 1139 1140 1141 1142 1143
			tx_desc->upper.data = 0;

			i++;
			if (i == tx_ring->count)
				i = 0;
		}

1144 1145
		if (i == tx_ring->next_to_use)
			break;
1146 1147 1148 1149 1150 1151
		eop = tx_ring->buffer_info[i].next_to_watch;
		eop_desc = E1000_TX_DESC(*tx_ring, eop);
	}

	tx_ring->next_to_clean = i;

1152 1153
	netdev_completed_queue(netdev, pkts_compl, bytes_compl);

1154
#define TX_WAKE_THRESHOLD 32
1155 1156
	if (count && netif_carrier_ok(netdev) &&
	    e1000_desc_unused(tx_ring) >= TX_WAKE_THRESHOLD) {
1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169
		/* Make sure that anybody stopping the queue after this
		 * sees the new next_to_clean.
		 */
		smp_mb();

		if (netif_queue_stopped(netdev) &&
		    !(test_bit(__E1000_DOWN, &adapter->state))) {
			netif_wake_queue(netdev);
			++adapter->restart_queue;
		}
	}

	if (adapter->detect_tx_hung) {
1170 1171 1172 1173
		/*
		 * Detect a transmit hang in hardware, this serializes the
		 * check with the clearing of time_stamp and movement of i
		 */
1174
		adapter->detect_tx_hung = false;
1175 1176
		if (tx_ring->buffer_info[i].time_stamp &&
		    time_after(jiffies, tx_ring->buffer_info[i].time_stamp
1177
			       + (adapter->tx_timeout_factor * HZ)) &&
1178
		    !(er32(STATUS) & E1000_STATUS_TXOFF))
1179
			schedule_work(&adapter->print_hang_task);
1180 1181
		else
			adapter->tx_hang_recheck = false;
1182 1183 1184
	}
	adapter->total_tx_bytes += total_tx_bytes;
	adapter->total_tx_packets += total_tx_packets;
1185
	return count < tx_ring->count;
1186 1187 1188 1189
}

/**
 * e1000_clean_rx_irq_ps - Send received data up the network stack; packet split
1190
 * @rx_ring: Rx descriptor ring
1191 1192 1193 1194
 *
 * the return value indicates whether actual cleaning was done, there
 * is no guarantee that everything was cleaned
 **/
1195 1196
static bool e1000_clean_rx_irq_ps(struct e1000_ring *rx_ring, int *work_done,
				  int work_to_do)
1197
{
1198
	struct e1000_adapter *adapter = rx_ring->adapter;
1199
	struct e1000_hw *hw = &adapter->hw;
1200 1201 1202 1203 1204 1205 1206 1207 1208
	union e1000_rx_desc_packet_split *rx_desc, *next_rxd;
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
	struct e1000_buffer *buffer_info, *next_buffer;
	struct e1000_ps_page *ps_page;
	struct sk_buff *skb;
	unsigned int i, j;
	u32 length, staterr;
	int cleaned_count = 0;
1209
	bool cleaned = false;
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221
	unsigned int total_rx_bytes = 0, total_rx_packets = 0;

	i = rx_ring->next_to_clean;
	rx_desc = E1000_RX_DESC_PS(*rx_ring, i);
	staterr = le32_to_cpu(rx_desc->wb.middle.status_error);
	buffer_info = &rx_ring->buffer_info[i];

	while (staterr & E1000_RXD_STAT_DD) {
		if (*work_done >= work_to_do)
			break;
		(*work_done)++;
		skb = buffer_info->skb;
1222
		rmb();	/* read descriptor and rx_buffer_info after status DD */
1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234

		/* in the packet split case this is header only */
		prefetch(skb->data - NET_IP_ALIGN);

		i++;
		if (i == rx_ring->count)
			i = 0;
		next_rxd = E1000_RX_DESC_PS(*rx_ring, i);
		prefetch(next_rxd);

		next_buffer = &rx_ring->buffer_info[i];

1235
		cleaned = true;
1236
		cleaned_count++;
1237
		dma_unmap_single(&pdev->dev, buffer_info->dma,
1238
				 adapter->rx_ps_bsize0, DMA_FROM_DEVICE);
1239 1240
		buffer_info->dma = 0;

1241
		/* see !EOP comment in other Rx routine */
1242 1243 1244 1245
		if (!(staterr & E1000_RXD_STAT_EOP))
			adapter->flags2 |= FLAG2_IS_DISCARDING;

		if (adapter->flags2 & FLAG2_IS_DISCARDING) {
1246
			e_dbg("Packet Split buffers didn't pick up the full packet\n");
1247
			dev_kfree_skb_irq(skb);
1248 1249
			if (staterr & E1000_RXD_STAT_EOP)
				adapter->flags2 &= ~FLAG2_IS_DISCARDING;
1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260
			goto next_desc;
		}

		if (staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) {
			dev_kfree_skb_irq(skb);
			goto next_desc;
		}

		length = le16_to_cpu(rx_desc->wb.middle.length0);

		if (!length) {
1261
			e_dbg("Last part of the packet spanning multiple descriptors\n");
1262 1263 1264 1265 1266 1267 1268 1269
			dev_kfree_skb_irq(skb);
			goto next_desc;
		}

		/* Good Receive */
		skb_put(skb, length);

		{
1270 1271 1272 1273
		/*
		 * this looks ugly, but it seems compiler issues make it
		 * more efficient than reusing j
		 */
1274 1275
		int l1 = le16_to_cpu(rx_desc->wb.upper.length[0]);

1276 1277 1278 1279 1280
		/*
		 * page alloc/put takes too long and effects small packet
		 * throughput, so unsplit small packets and save the alloc/put
		 * only valid in softirq (napi) context to call kmap_*
		 */
1281 1282 1283 1284
		if (l1 && (l1 <= copybreak) &&
		    ((length + l1) <= adapter->rx_ps_bsize0)) {
			u8 *vaddr;

A
Auke Kok 已提交
1285
			ps_page = &buffer_info->ps_pages[0];
1286

1287 1288
			/*
			 * there is no documentation about how to call
1289
			 * kmap_atomic, so we can't hold the mapping
1290 1291
			 * very long
			 */
1292 1293
			dma_sync_single_for_cpu(&pdev->dev, ps_page->dma,
						PAGE_SIZE, DMA_FROM_DEVICE);
1294 1295 1296
			vaddr = kmap_atomic(ps_page->page, KM_SKB_DATA_SOFTIRQ);
			memcpy(skb_tail_pointer(skb), vaddr, l1);
			kunmap_atomic(vaddr, KM_SKB_DATA_SOFTIRQ);
1297 1298
			dma_sync_single_for_device(&pdev->dev, ps_page->dma,
						   PAGE_SIZE, DMA_FROM_DEVICE);
A
Auke Kok 已提交
1299

J
Jeff Kirsher 已提交
1300 1301 1302 1303
			/* remove the CRC */
			if (!(adapter->flags2 & FLAG2_CRC_STRIPPING))
				l1 -= 4;

1304 1305 1306 1307 1308 1309 1310 1311 1312 1313
			skb_put(skb, l1);
			goto copydone;
		} /* if */
		}

		for (j = 0; j < PS_PAGE_BUFFERS; j++) {
			length = le16_to_cpu(rx_desc->wb.upper.length[j]);
			if (!length)
				break;

A
Auke Kok 已提交
1314
			ps_page = &buffer_info->ps_pages[j];
1315 1316
			dma_unmap_page(&pdev->dev, ps_page->dma, PAGE_SIZE,
				       DMA_FROM_DEVICE);
1317 1318 1319 1320 1321
			ps_page->dma = 0;
			skb_fill_page_desc(skb, j, ps_page->page, 0, length);
			ps_page->page = NULL;
			skb->len += length;
			skb->data_len += length;
1322
			skb->truesize += PAGE_SIZE;
1323 1324
		}

J
Jeff Kirsher 已提交
1325 1326 1327 1328 1329 1330
		/* strip the ethernet crc, problem is we're using pages now so
		 * this whole operation can get a little cpu intensive
		 */
		if (!(adapter->flags2 & FLAG2_CRC_STRIPPING))
			pskb_trim(skb, skb->len - 4);

1331 1332 1333 1334
copydone:
		total_rx_bytes += skb->len;
		total_rx_packets++;

1335 1336
		e1000_rx_checksum(adapter, staterr,
				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
1337

1338 1339
		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);

1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352
		if (rx_desc->wb.upper.header_status &
			   cpu_to_le16(E1000_RXDPS_HDRSTAT_HDRSP))
			adapter->rx_hdr_split++;

		e1000_receive_skb(adapter, netdev, skb,
				  staterr, rx_desc->wb.middle.vlan);

next_desc:
		rx_desc->wb.middle.status_error &= cpu_to_le32(~0xFF);
		buffer_info->skb = NULL;

		/* return some buffers to hardware, one at a time is too slow */
		if (cleaned_count >= E1000_RX_BUFFER_WRITE) {
1353
			adapter->alloc_rx_buf(rx_ring, cleaned_count,
1354
					      GFP_ATOMIC);
1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367
			cleaned_count = 0;
		}

		/* use prefetched values */
		rx_desc = next_rxd;
		buffer_info = next_buffer;

		staterr = le32_to_cpu(rx_desc->wb.middle.status_error);
	}
	rx_ring->next_to_clean = i;

	cleaned_count = e1000_desc_unused(rx_ring);
	if (cleaned_count)
1368
		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
1369 1370

	adapter->total_rx_bytes += total_rx_bytes;
1371
	adapter->total_rx_packets += total_rx_packets;
1372 1373 1374
	return cleaned;
}

1375 1376 1377 1378 1379 1380 1381 1382 1383
/**
 * e1000_consume_page - helper function
 **/
static void e1000_consume_page(struct e1000_buffer *bi, struct sk_buff *skb,
                               u16 length)
{
	bi->page = NULL;
	skb->len += length;
	skb->data_len += length;
1384
	skb->truesize += PAGE_SIZE;
1385 1386 1387 1388 1389 1390 1391 1392 1393
}

/**
 * e1000_clean_jumbo_rx_irq - Send received data up the network stack; legacy
 * @adapter: board private structure
 *
 * the return value indicates whether actual cleaning was done, there
 * is no guarantee that everything was cleaned
 **/
1394 1395
static bool e1000_clean_jumbo_rx_irq(struct e1000_ring *rx_ring, int *work_done,
				     int work_to_do)
1396
{
1397
	struct e1000_adapter *adapter = rx_ring->adapter;
1398 1399
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
1400
	union e1000_rx_desc_extended *rx_desc, *next_rxd;
1401
	struct e1000_buffer *buffer_info, *next_buffer;
1402
	u32 length, staterr;
1403 1404 1405 1406 1407 1408
	unsigned int i;
	int cleaned_count = 0;
	bool cleaned = false;
	unsigned int total_rx_bytes=0, total_rx_packets=0;

	i = rx_ring->next_to_clean;
1409 1410
	rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
	staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
1411 1412
	buffer_info = &rx_ring->buffer_info[i];

1413
	while (staterr & E1000_RXD_STAT_DD) {
1414 1415 1416 1417 1418
		struct sk_buff *skb;

		if (*work_done >= work_to_do)
			break;
		(*work_done)++;
1419
		rmb();	/* read descriptor and rx_buffer_info after status DD */
1420 1421 1422 1423 1424 1425 1426

		skb = buffer_info->skb;
		buffer_info->skb = NULL;

		++i;
		if (i == rx_ring->count)
			i = 0;
1427
		next_rxd = E1000_RX_DESC_EXT(*rx_ring, i);
1428 1429 1430 1431 1432 1433
		prefetch(next_rxd);

		next_buffer = &rx_ring->buffer_info[i];

		cleaned = true;
		cleaned_count++;
1434 1435
		dma_unmap_page(&pdev->dev, buffer_info->dma, PAGE_SIZE,
			       DMA_FROM_DEVICE);
1436 1437
		buffer_info->dma = 0;

1438
		length = le16_to_cpu(rx_desc->wb.upper.length);
1439 1440

		/* errors is only valid for DD + EOP descriptors */
1441 1442 1443 1444 1445 1446 1447 1448 1449
		if (unlikely((staterr & E1000_RXD_STAT_EOP) &&
			     (staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK))) {
			/* recycle both page and skb */
			buffer_info->skb = skb;
			/* an error means any chain goes out the window too */
			if (rx_ring->rx_skb_top)
				dev_kfree_skb_irq(rx_ring->rx_skb_top);
			rx_ring->rx_skb_top = NULL;
			goto next_desc;
1450 1451
		}

1452
#define rxtop (rx_ring->rx_skb_top)
1453
		if (!(staterr & E1000_RXD_STAT_EOP)) {
1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507
			/* this descriptor is only the beginning (or middle) */
			if (!rxtop) {
				/* this is the beginning of a chain */
				rxtop = skb;
				skb_fill_page_desc(rxtop, 0, buffer_info->page,
				                   0, length);
			} else {
				/* this is the middle of a chain */
				skb_fill_page_desc(rxtop,
				    skb_shinfo(rxtop)->nr_frags,
				    buffer_info->page, 0, length);
				/* re-use the skb, only consumed the page */
				buffer_info->skb = skb;
			}
			e1000_consume_page(buffer_info, rxtop, length);
			goto next_desc;
		} else {
			if (rxtop) {
				/* end of the chain */
				skb_fill_page_desc(rxtop,
				    skb_shinfo(rxtop)->nr_frags,
				    buffer_info->page, 0, length);
				/* re-use the current skb, we only consumed the
				 * page */
				buffer_info->skb = skb;
				skb = rxtop;
				rxtop = NULL;
				e1000_consume_page(buffer_info, skb, length);
			} else {
				/* no chain, got EOP, this buf is the packet
				 * copybreak to save the put_page/alloc_page */
				if (length <= copybreak &&
				    skb_tailroom(skb) >= length) {
					u8 *vaddr;
					vaddr = kmap_atomic(buffer_info->page,
					                   KM_SKB_DATA_SOFTIRQ);
					memcpy(skb_tail_pointer(skb), vaddr,
					       length);
					kunmap_atomic(vaddr,
					              KM_SKB_DATA_SOFTIRQ);
					/* re-use the page, so don't erase
					 * buffer_info->page */
					skb_put(skb, length);
				} else {
					skb_fill_page_desc(skb, 0,
					                   buffer_info->page, 0,
				                           length);
					e1000_consume_page(buffer_info, skb,
					                   length);
				}
			}
		}

		/* Receive Checksum Offload XXX recompute due to CRC strip? */
1508
		e1000_rx_checksum(adapter, staterr,
1509
				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
1510

1511 1512
		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);

1513 1514 1515 1516 1517 1518
		/* probably a little skewed due to removing CRC */
		total_rx_bytes += skb->len;
		total_rx_packets++;

		/* eth type trans needs skb->data to point to something */
		if (!pskb_may_pull(skb, ETH_HLEN)) {
1519
			e_err("pskb_may_pull failed.\n");
1520
			dev_kfree_skb_irq(skb);
1521 1522 1523
			goto next_desc;
		}

1524 1525
		e1000_receive_skb(adapter, netdev, skb, staterr,
				  rx_desc->wb.upper.vlan);
1526 1527

next_desc:
1528
		rx_desc->wb.upper.status_error &= cpu_to_le32(~0xFF);
1529 1530 1531

		/* return some buffers to hardware, one at a time is too slow */
		if (unlikely(cleaned_count >= E1000_RX_BUFFER_WRITE)) {
1532
			adapter->alloc_rx_buf(rx_ring, cleaned_count,
1533
					      GFP_ATOMIC);
1534 1535 1536 1537 1538 1539
			cleaned_count = 0;
		}

		/* use prefetched values */
		rx_desc = next_rxd;
		buffer_info = next_buffer;
1540 1541

		staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
1542 1543 1544 1545 1546
	}
	rx_ring->next_to_clean = i;

	cleaned_count = e1000_desc_unused(rx_ring);
	if (cleaned_count)
1547
		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
1548 1549 1550 1551 1552 1553

	adapter->total_rx_bytes += total_rx_bytes;
	adapter->total_rx_packets += total_rx_packets;
	return cleaned;
}

1554 1555
/**
 * e1000_clean_rx_ring - Free Rx Buffers per Queue
1556
 * @rx_ring: Rx descriptor ring
1557
 **/
1558
static void e1000_clean_rx_ring(struct e1000_ring *rx_ring)
1559
{
1560
	struct e1000_adapter *adapter = rx_ring->adapter;
1561 1562 1563 1564 1565 1566 1567 1568 1569 1570
	struct e1000_buffer *buffer_info;
	struct e1000_ps_page *ps_page;
	struct pci_dev *pdev = adapter->pdev;
	unsigned int i, j;

	/* Free all the Rx ring sk_buffs */
	for (i = 0; i < rx_ring->count; i++) {
		buffer_info = &rx_ring->buffer_info[i];
		if (buffer_info->dma) {
			if (adapter->clean_rx == e1000_clean_rx_irq)
1571
				dma_unmap_single(&pdev->dev, buffer_info->dma,
1572
						 adapter->rx_buffer_len,
1573
						 DMA_FROM_DEVICE);
1574
			else if (adapter->clean_rx == e1000_clean_jumbo_rx_irq)
1575
				dma_unmap_page(&pdev->dev, buffer_info->dma,
1576
				               PAGE_SIZE,
1577
					       DMA_FROM_DEVICE);
1578
			else if (adapter->clean_rx == e1000_clean_rx_irq_ps)
1579
				dma_unmap_single(&pdev->dev, buffer_info->dma,
1580
						 adapter->rx_ps_bsize0,
1581
						 DMA_FROM_DEVICE);
1582 1583 1584
			buffer_info->dma = 0;
		}

1585 1586 1587 1588 1589
		if (buffer_info->page) {
			put_page(buffer_info->page);
			buffer_info->page = NULL;
		}

1590 1591 1592 1593 1594 1595
		if (buffer_info->skb) {
			dev_kfree_skb(buffer_info->skb);
			buffer_info->skb = NULL;
		}

		for (j = 0; j < PS_PAGE_BUFFERS; j++) {
A
Auke Kok 已提交
1596
			ps_page = &buffer_info->ps_pages[j];
1597 1598
			if (!ps_page->page)
				break;
1599 1600
			dma_unmap_page(&pdev->dev, ps_page->dma, PAGE_SIZE,
				       DMA_FROM_DEVICE);
1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617
			ps_page->dma = 0;
			put_page(ps_page->page);
			ps_page->page = NULL;
		}
	}

	/* there also may be some cached data from a chained receive */
	if (rx_ring->rx_skb_top) {
		dev_kfree_skb(rx_ring->rx_skb_top);
		rx_ring->rx_skb_top = NULL;
	}

	/* Zero out the descriptor ring */
	memset(rx_ring->desc, 0, rx_ring->size);

	rx_ring->next_to_clean = 0;
	rx_ring->next_to_use = 0;
1618
	adapter->flags2 &= ~FLAG2_IS_DISCARDING;
1619

1620 1621
	writel(0, rx_ring->head);
	writel(0, rx_ring->tail);
1622 1623
}

1624 1625 1626 1627 1628
static void e1000e_downshift_workaround(struct work_struct *work)
{
	struct e1000_adapter *adapter = container_of(work,
					struct e1000_adapter, downshift_task);

1629 1630 1631
	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

1632 1633 1634
	e1000e_gig_downshift_workaround_ich8lan(&adapter->hw);
}

1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646
/**
 * e1000_intr_msi - Interrupt Handler
 * @irq: interrupt number
 * @data: pointer to a network interface device structure
 **/
static irqreturn_t e1000_intr_msi(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 icr = er32(ICR);

1647 1648 1649
	/*
	 * read ICR disables interrupts using IAM
	 */
1650

1651
	if (icr & E1000_ICR_LSC) {
1652
		hw->mac.get_link_status = 1;
1653 1654 1655 1656
		/*
		 * ICH8 workaround-- Call gig speed drop workaround on cable
		 * disconnect (LSC) before accessing any PHY registers
		 */
1657 1658
		if ((adapter->flags & FLAG_LSC_GIG_SPEED_DROP) &&
		    (!(er32(STATUS) & E1000_STATUS_LU)))
1659
			schedule_work(&adapter->downshift_task);
1660

1661 1662
		/*
		 * 80003ES2LAN workaround-- For packet buffer work-around on
1663
		 * link down event; disable receives here in the ISR and reset
1664 1665
		 * adapter in watchdog
		 */
1666 1667 1668 1669 1670
		if (netif_carrier_ok(netdev) &&
		    adapter->flags & FLAG_RX_NEEDS_RESTART) {
			/* disable receives */
			u32 rctl = er32(RCTL);
			ew32(RCTL, rctl & ~E1000_RCTL_EN);
1671
			adapter->flags |= FLAG_RX_RESTART_NOW;
1672 1673 1674 1675 1676 1677
		}
		/* guard against interrupt when we're going down */
		if (!test_bit(__E1000_DOWN, &adapter->state))
			mod_timer(&adapter->watchdog_timer, jiffies + 1);
	}

1678
	if (napi_schedule_prep(&adapter->napi)) {
1679 1680 1681 1682
		adapter->total_tx_bytes = 0;
		adapter->total_tx_packets = 0;
		adapter->total_rx_bytes = 0;
		adapter->total_rx_packets = 0;
1683
		__napi_schedule(&adapter->napi);
1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699
	}

	return IRQ_HANDLED;
}

/**
 * e1000_intr - Interrupt Handler
 * @irq: interrupt number
 * @data: pointer to a network interface device structure
 **/
static irqreturn_t e1000_intr(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl, icr = er32(ICR);
1700

1701
	if (!icr || test_bit(__E1000_DOWN, &adapter->state))
1702 1703
		return IRQ_NONE;  /* Not our interrupt */

1704 1705 1706 1707
	/*
	 * IMS will not auto-mask if INT_ASSERTED is not set, and if it is
	 * not set, then the adapter didn't send an interrupt
	 */
1708 1709 1710
	if (!(icr & E1000_ICR_INT_ASSERTED))
		return IRQ_NONE;

1711 1712 1713 1714 1715
	/*
	 * Interrupt Auto-Mask...upon reading ICR,
	 * interrupts are masked.  No need for the
	 * IMC write
	 */
1716

1717
	if (icr & E1000_ICR_LSC) {
1718
		hw->mac.get_link_status = 1;
1719 1720 1721 1722
		/*
		 * ICH8 workaround-- Call gig speed drop workaround on cable
		 * disconnect (LSC) before accessing any PHY registers
		 */
1723 1724
		if ((adapter->flags & FLAG_LSC_GIG_SPEED_DROP) &&
		    (!(er32(STATUS) & E1000_STATUS_LU)))
1725
			schedule_work(&adapter->downshift_task);
1726

1727 1728
		/*
		 * 80003ES2LAN workaround--
1729 1730 1731 1732 1733 1734 1735 1736 1737
		 * For packet buffer work-around on link down event;
		 * disable receives here in the ISR and
		 * reset adapter in watchdog
		 */
		if (netif_carrier_ok(netdev) &&
		    (adapter->flags & FLAG_RX_NEEDS_RESTART)) {
			/* disable receives */
			rctl = er32(RCTL);
			ew32(RCTL, rctl & ~E1000_RCTL_EN);
1738
			adapter->flags |= FLAG_RX_RESTART_NOW;
1739 1740 1741 1742 1743 1744
		}
		/* guard against interrupt when we're going down */
		if (!test_bit(__E1000_DOWN, &adapter->state))
			mod_timer(&adapter->watchdog_timer, jiffies + 1);
	}

1745
	if (napi_schedule_prep(&adapter->napi)) {
1746 1747 1748 1749
		adapter->total_tx_bytes = 0;
		adapter->total_tx_packets = 0;
		adapter->total_rx_bytes = 0;
		adapter->total_rx_packets = 0;
1750
		__napi_schedule(&adapter->napi);
1751 1752 1753 1754 1755
	}

	return IRQ_HANDLED;
}

1756 1757 1758 1759 1760 1761 1762 1763
static irqreturn_t e1000_msix_other(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 icr = er32(ICR);

	if (!(icr & E1000_ICR_INT_ASSERTED)) {
1764 1765
		if (!test_bit(__E1000_DOWN, &adapter->state))
			ew32(IMS, E1000_IMS_OTHER);
1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781
		return IRQ_NONE;
	}

	if (icr & adapter->eiac_mask)
		ew32(ICS, (icr & adapter->eiac_mask));

	if (icr & E1000_ICR_OTHER) {
		if (!(icr & E1000_ICR_LSC))
			goto no_link_interrupt;
		hw->mac.get_link_status = 1;
		/* guard against interrupt when we're going down */
		if (!test_bit(__E1000_DOWN, &adapter->state))
			mod_timer(&adapter->watchdog_timer, jiffies + 1);
	}

no_link_interrupt:
1782 1783
	if (!test_bit(__E1000_DOWN, &adapter->state))
		ew32(IMS, E1000_IMS_LSC | E1000_IMS_OTHER);
1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799

	return IRQ_HANDLED;
}


static irqreturn_t e1000_intr_msix_tx(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_ring *tx_ring = adapter->tx_ring;


	adapter->total_tx_bytes = 0;
	adapter->total_tx_packets = 0;

1800
	if (!e1000_clean_tx_irq(tx_ring))
1801 1802 1803 1804 1805 1806 1807 1808 1809 1810
		/* Ring was not completely cleaned, so fire another interrupt */
		ew32(ICS, tx_ring->ims_val);

	return IRQ_HANDLED;
}

static irqreturn_t e1000_intr_msix_rx(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
1811
	struct e1000_ring *rx_ring = adapter->rx_ring;
1812 1813 1814 1815

	/* Write the ITR value calculated at the end of the
	 * previous interrupt.
	 */
1816 1817 1818 1819
	if (rx_ring->set_itr) {
		writel(1000000000 / (rx_ring->itr_val * 256),
		       rx_ring->itr_register);
		rx_ring->set_itr = 0;
1820 1821
	}

1822
	if (napi_schedule_prep(&adapter->napi)) {
1823 1824
		adapter->total_rx_bytes = 0;
		adapter->total_rx_packets = 0;
1825
		__napi_schedule(&adapter->napi);
1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858
	}
	return IRQ_HANDLED;
}

/**
 * e1000_configure_msix - Configure MSI-X hardware
 *
 * e1000_configure_msix sets up the hardware to properly
 * generate MSI-X interrupts.
 **/
static void e1000_configure_msix(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_ring *rx_ring = adapter->rx_ring;
	struct e1000_ring *tx_ring = adapter->tx_ring;
	int vector = 0;
	u32 ctrl_ext, ivar = 0;

	adapter->eiac_mask = 0;

	/* Workaround issue with spurious interrupts on 82574 in MSI-X mode */
	if (hw->mac.type == e1000_82574) {
		u32 rfctl = er32(RFCTL);
		rfctl |= E1000_RFCTL_ACK_DIS;
		ew32(RFCTL, rfctl);
	}

#define E1000_IVAR_INT_ALLOC_VALID	0x8
	/* Configure Rx vector */
	rx_ring->ims_val = E1000_IMS_RXQ0;
	adapter->eiac_mask |= rx_ring->ims_val;
	if (rx_ring->itr_val)
		writel(1000000000 / (rx_ring->itr_val * 256),
1859
		       rx_ring->itr_register);
1860
	else
1861
		writel(1, rx_ring->itr_register);
1862 1863 1864 1865 1866 1867 1868
	ivar = E1000_IVAR_INT_ALLOC_VALID | vector;

	/* Configure Tx vector */
	tx_ring->ims_val = E1000_IMS_TXQ0;
	vector++;
	if (tx_ring->itr_val)
		writel(1000000000 / (tx_ring->itr_val * 256),
1869
		       tx_ring->itr_register);
1870
	else
1871
		writel(1, tx_ring->itr_register);
1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921
	adapter->eiac_mask |= tx_ring->ims_val;
	ivar |= ((E1000_IVAR_INT_ALLOC_VALID | vector) << 8);

	/* set vector for Other Causes, e.g. link changes */
	vector++;
	ivar |= ((E1000_IVAR_INT_ALLOC_VALID | vector) << 16);
	if (rx_ring->itr_val)
		writel(1000000000 / (rx_ring->itr_val * 256),
		       hw->hw_addr + E1000_EITR_82574(vector));
	else
		writel(1, hw->hw_addr + E1000_EITR_82574(vector));

	/* Cause Tx interrupts on every write back */
	ivar |= (1 << 31);

	ew32(IVAR, ivar);

	/* enable MSI-X PBA support */
	ctrl_ext = er32(CTRL_EXT);
	ctrl_ext |= E1000_CTRL_EXT_PBA_CLR;

	/* Auto-Mask Other interrupts upon ICR read */
#define E1000_EIAC_MASK_82574   0x01F00000
	ew32(IAM, ~E1000_EIAC_MASK_82574 | E1000_IMS_OTHER);
	ctrl_ext |= E1000_CTRL_EXT_EIAME;
	ew32(CTRL_EXT, ctrl_ext);
	e1e_flush();
}

void e1000e_reset_interrupt_capability(struct e1000_adapter *adapter)
{
	if (adapter->msix_entries) {
		pci_disable_msix(adapter->pdev);
		kfree(adapter->msix_entries);
		adapter->msix_entries = NULL;
	} else if (adapter->flags & FLAG_MSI_ENABLED) {
		pci_disable_msi(adapter->pdev);
		adapter->flags &= ~FLAG_MSI_ENABLED;
	}
}

/**
 * e1000e_set_interrupt_capability - set MSI or MSI-X if supported
 *
 * Attempt to configure interrupts using the best available
 * capabilities of the hardware and kernel.
 **/
void e1000e_set_interrupt_capability(struct e1000_adapter *adapter)
{
	int err;
1922
	int i;
1923 1924 1925 1926

	switch (adapter->int_mode) {
	case E1000E_INT_MODE_MSIX:
		if (adapter->flags & FLAG_HAS_MSIX) {
1927 1928
			adapter->num_vectors = 3; /* RxQ0, TxQ0 and other */
			adapter->msix_entries = kcalloc(adapter->num_vectors,
1929 1930 1931
						      sizeof(struct msix_entry),
						      GFP_KERNEL);
			if (adapter->msix_entries) {
1932
				for (i = 0; i < adapter->num_vectors; i++)
1933 1934 1935 1936
					adapter->msix_entries[i].entry = i;

				err = pci_enable_msix(adapter->pdev,
						      adapter->msix_entries,
1937
						      adapter->num_vectors);
B
Bruce Allan 已提交
1938
				if (err == 0)
1939 1940 1941
					return;
			}
			/* MSI-X failed, so fall through and try MSI */
1942
			e_err("Failed to initialize MSI-X interrupts.  Falling back to MSI interrupts.\n");
1943 1944 1945 1946 1947 1948 1949 1950 1951
			e1000e_reset_interrupt_capability(adapter);
		}
		adapter->int_mode = E1000E_INT_MODE_MSI;
		/* Fall through */
	case E1000E_INT_MODE_MSI:
		if (!pci_enable_msi(adapter->pdev)) {
			adapter->flags |= FLAG_MSI_ENABLED;
		} else {
			adapter->int_mode = E1000E_INT_MODE_LEGACY;
1952
			e_err("Failed to initialize MSI interrupts.  Falling back to legacy interrupts.\n");
1953 1954 1955 1956 1957 1958
		}
		/* Fall through */
	case E1000E_INT_MODE_LEGACY:
		/* Don't do anything; this is the system default */
		break;
	}
1959 1960 1961

	/* store the number of vectors being used */
	adapter->num_vectors = 1;
1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975
}

/**
 * e1000_request_msix - Initialize MSI-X interrupts
 *
 * e1000_request_msix allocates MSI-X vectors and requests interrupts from the
 * kernel.
 **/
static int e1000_request_msix(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	int err = 0, vector = 0;

	if (strlen(netdev->name) < (IFNAMSIZ - 5))
1976 1977 1978
		snprintf(adapter->rx_ring->name,
			 sizeof(adapter->rx_ring->name) - 1,
			 "%s-rx-0", netdev->name);
1979 1980 1981
	else
		memcpy(adapter->rx_ring->name, netdev->name, IFNAMSIZ);
	err = request_irq(adapter->msix_entries[vector].vector,
1982
			  e1000_intr_msix_rx, 0, adapter->rx_ring->name,
1983 1984 1985
			  netdev);
	if (err)
		goto out;
1986 1987
	adapter->rx_ring->itr_register = adapter->hw.hw_addr +
	    E1000_EITR_82574(vector);
1988 1989 1990 1991
	adapter->rx_ring->itr_val = adapter->itr;
	vector++;

	if (strlen(netdev->name) < (IFNAMSIZ - 5))
1992 1993 1994
		snprintf(adapter->tx_ring->name,
			 sizeof(adapter->tx_ring->name) - 1,
			 "%s-tx-0", netdev->name);
1995 1996 1997
	else
		memcpy(adapter->tx_ring->name, netdev->name, IFNAMSIZ);
	err = request_irq(adapter->msix_entries[vector].vector,
1998
			  e1000_intr_msix_tx, 0, adapter->tx_ring->name,
1999 2000 2001
			  netdev);
	if (err)
		goto out;
2002 2003
	adapter->tx_ring->itr_register = adapter->hw.hw_addr +
	    E1000_EITR_82574(vector);
2004 2005 2006 2007
	adapter->tx_ring->itr_val = adapter->itr;
	vector++;

	err = request_irq(adapter->msix_entries[vector].vector,
2008
			  e1000_msix_other, 0, netdev->name, netdev);
2009 2010 2011 2012 2013 2014 2015 2016 2017
	if (err)
		goto out;

	e1000_configure_msix(adapter);
	return 0;
out:
	return err;
}

2018 2019 2020 2021 2022 2023
/**
 * e1000_request_irq - initialize interrupts
 *
 * Attempts to configure interrupts using the best available
 * capabilities of the hardware and kernel.
 **/
2024 2025 2026 2027 2028
static int e1000_request_irq(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	int err;

2029 2030 2031 2032 2033 2034 2035 2036
	if (adapter->msix_entries) {
		err = e1000_request_msix(adapter);
		if (!err)
			return err;
		/* fall back to MSI */
		e1000e_reset_interrupt_capability(adapter);
		adapter->int_mode = E1000E_INT_MODE_MSI;
		e1000e_set_interrupt_capability(adapter);
2037
	}
2038
	if (adapter->flags & FLAG_MSI_ENABLED) {
2039
		err = request_irq(adapter->pdev->irq, e1000_intr_msi, 0,
2040 2041 2042
				  netdev->name, netdev);
		if (!err)
			return err;
2043

2044 2045 2046
		/* fall back to legacy interrupt */
		e1000e_reset_interrupt_capability(adapter);
		adapter->int_mode = E1000E_INT_MODE_LEGACY;
2047 2048
	}

2049
	err = request_irq(adapter->pdev->irq, e1000_intr, IRQF_SHARED,
2050 2051 2052 2053
			  netdev->name, netdev);
	if (err)
		e_err("Unable to allocate interrupt, Error: %d\n", err);

2054 2055 2056 2057 2058 2059 2060
	return err;
}

static void e1000_free_irq(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;

2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072
	if (adapter->msix_entries) {
		int vector = 0;

		free_irq(adapter->msix_entries[vector].vector, netdev);
		vector++;

		free_irq(adapter->msix_entries[vector].vector, netdev);
		vector++;

		/* Other Causes interrupt vector */
		free_irq(adapter->msix_entries[vector].vector, netdev);
		return;
2073
	}
2074 2075

	free_irq(adapter->pdev->irq, netdev);
2076 2077 2078 2079 2080 2081 2082 2083 2084 2085
}

/**
 * e1000_irq_disable - Mask off interrupt generation on the NIC
 **/
static void e1000_irq_disable(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

	ew32(IMC, ~0);
2086 2087
	if (adapter->msix_entries)
		ew32(EIAC_82574, 0);
2088
	e1e_flush();
2089 2090 2091 2092 2093 2094 2095 2096

	if (adapter->msix_entries) {
		int i;
		for (i = 0; i < adapter->num_vectors; i++)
			synchronize_irq(adapter->msix_entries[i].vector);
	} else {
		synchronize_irq(adapter->pdev->irq);
	}
2097 2098 2099 2100 2101 2102 2103 2104 2105
}

/**
 * e1000_irq_enable - Enable default interrupt generation settings
 **/
static void e1000_irq_enable(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

2106 2107 2108 2109 2110 2111
	if (adapter->msix_entries) {
		ew32(EIAC_82574, adapter->eiac_mask & E1000_EIAC_MASK_82574);
		ew32(IMS, adapter->eiac_mask | E1000_IMS_OTHER | E1000_IMS_LSC);
	} else {
		ew32(IMS, IMS_ENABLE_MASK);
	}
J
Jesse Brandeburg 已提交
2112
	e1e_flush();
2113 2114 2115
}

/**
2116
 * e1000e_get_hw_control - get control of the h/w from f/w
2117 2118
 * @adapter: address of board private structure
 *
2119
 * e1000e_get_hw_control sets {CTRL_EXT|SWSM}:DRV_LOAD bit.
2120 2121 2122 2123
 * For ASF and Pass Through versions of f/w this means that
 * the driver is loaded. For AMT version (only with 82573)
 * of the f/w this means that the network i/f is open.
 **/
2124
void e1000e_get_hw_control(struct e1000_adapter *adapter)
2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135
{
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl_ext;
	u32 swsm;

	/* Let firmware know the driver has taken over */
	if (adapter->flags & FLAG_HAS_SWSM_ON_LOAD) {
		swsm = er32(SWSM);
		ew32(SWSM, swsm | E1000_SWSM_DRV_LOAD);
	} else if (adapter->flags & FLAG_HAS_CTRLEXT_ON_LOAD) {
		ctrl_ext = er32(CTRL_EXT);
2136
		ew32(CTRL_EXT, ctrl_ext | E1000_CTRL_EXT_DRV_LOAD);
2137 2138 2139 2140
	}
}

/**
2141
 * e1000e_release_hw_control - release control of the h/w to f/w
2142 2143
 * @adapter: address of board private structure
 *
2144
 * e1000e_release_hw_control resets {CTRL_EXT|SWSM}:DRV_LOAD bit.
2145 2146 2147 2148 2149
 * For ASF and Pass Through versions of f/w this means that the
 * driver is no longer loaded. For AMT version (only with 82573) i
 * of the f/w this means that the network i/f is closed.
 *
 **/
2150
void e1000e_release_hw_control(struct e1000_adapter *adapter)
2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161
{
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl_ext;
	u32 swsm;

	/* Let firmware taken over control of h/w */
	if (adapter->flags & FLAG_HAS_SWSM_ON_LOAD) {
		swsm = er32(SWSM);
		ew32(SWSM, swsm & ~E1000_SWSM_DRV_LOAD);
	} else if (adapter->flags & FLAG_HAS_CTRLEXT_ON_LOAD) {
		ctrl_ext = er32(CTRL_EXT);
2162
		ew32(CTRL_EXT, ctrl_ext & ~E1000_CTRL_EXT_DRV_LOAD);
2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183
	}
}

/**
 * @e1000_alloc_ring - allocate memory for a ring structure
 **/
static int e1000_alloc_ring_dma(struct e1000_adapter *adapter,
				struct e1000_ring *ring)
{
	struct pci_dev *pdev = adapter->pdev;

	ring->desc = dma_alloc_coherent(&pdev->dev, ring->size, &ring->dma,
					GFP_KERNEL);
	if (!ring->desc)
		return -ENOMEM;

	return 0;
}

/**
 * e1000e_setup_tx_resources - allocate Tx resources (Descriptors)
2184
 * @tx_ring: Tx descriptor ring
2185 2186 2187
 *
 * Return 0 on success, negative on failure
 **/
2188
int e1000e_setup_tx_resources(struct e1000_ring *tx_ring)
2189
{
2190
	struct e1000_adapter *adapter = tx_ring->adapter;
2191 2192 2193
	int err = -ENOMEM, size;

	size = sizeof(struct e1000_buffer) * tx_ring->count;
E
Eric Dumazet 已提交
2194
	tx_ring->buffer_info = vzalloc(size);
2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211
	if (!tx_ring->buffer_info)
		goto err;

	/* round up to nearest 4K */
	tx_ring->size = tx_ring->count * sizeof(struct e1000_tx_desc);
	tx_ring->size = ALIGN(tx_ring->size, 4096);

	err = e1000_alloc_ring_dma(adapter, tx_ring);
	if (err)
		goto err;

	tx_ring->next_to_use = 0;
	tx_ring->next_to_clean = 0;

	return 0;
err:
	vfree(tx_ring->buffer_info);
2212
	e_err("Unable to allocate memory for the transmit descriptor ring\n");
2213 2214 2215 2216 2217
	return err;
}

/**
 * e1000e_setup_rx_resources - allocate Rx resources (Descriptors)
2218
 * @rx_ring: Rx descriptor ring
2219 2220 2221
 *
 * Returns 0 on success, negative on failure
 **/
2222
int e1000e_setup_rx_resources(struct e1000_ring *rx_ring)
2223
{
2224
	struct e1000_adapter *adapter = rx_ring->adapter;
A
Auke Kok 已提交
2225 2226
	struct e1000_buffer *buffer_info;
	int i, size, desc_len, err = -ENOMEM;
2227 2228

	size = sizeof(struct e1000_buffer) * rx_ring->count;
E
Eric Dumazet 已提交
2229
	rx_ring->buffer_info = vzalloc(size);
2230 2231 2232
	if (!rx_ring->buffer_info)
		goto err;

A
Auke Kok 已提交
2233 2234 2235 2236 2237 2238 2239 2240
	for (i = 0; i < rx_ring->count; i++) {
		buffer_info = &rx_ring->buffer_info[i];
		buffer_info->ps_pages = kcalloc(PS_PAGE_BUFFERS,
						sizeof(struct e1000_ps_page),
						GFP_KERNEL);
		if (!buffer_info->ps_pages)
			goto err_pages;
	}
2241 2242 2243 2244 2245 2246 2247 2248 2249

	desc_len = sizeof(union e1000_rx_desc_packet_split);

	/* Round up to nearest 4K */
	rx_ring->size = rx_ring->count * desc_len;
	rx_ring->size = ALIGN(rx_ring->size, 4096);

	err = e1000_alloc_ring_dma(adapter, rx_ring);
	if (err)
A
Auke Kok 已提交
2250
		goto err_pages;
2251 2252 2253 2254 2255 2256

	rx_ring->next_to_clean = 0;
	rx_ring->next_to_use = 0;
	rx_ring->rx_skb_top = NULL;

	return 0;
A
Auke Kok 已提交
2257 2258 2259 2260 2261 2262

err_pages:
	for (i = 0; i < rx_ring->count; i++) {
		buffer_info = &rx_ring->buffer_info[i];
		kfree(buffer_info->ps_pages);
	}
2263 2264
err:
	vfree(rx_ring->buffer_info);
2265
	e_err("Unable to allocate memory for the receive descriptor ring\n");
2266 2267 2268 2269 2270
	return err;
}

/**
 * e1000_clean_tx_ring - Free Tx Buffers
2271
 * @tx_ring: Tx descriptor ring
2272
 **/
2273
static void e1000_clean_tx_ring(struct e1000_ring *tx_ring)
2274
{
2275
	struct e1000_adapter *adapter = tx_ring->adapter;
2276 2277 2278 2279 2280 2281
	struct e1000_buffer *buffer_info;
	unsigned long size;
	unsigned int i;

	for (i = 0; i < tx_ring->count; i++) {
		buffer_info = &tx_ring->buffer_info[i];
2282
		e1000_put_txbuf(tx_ring, buffer_info);
2283 2284
	}

2285
	netdev_reset_queue(adapter->netdev);
2286 2287 2288 2289 2290 2291 2292 2293
	size = sizeof(struct e1000_buffer) * tx_ring->count;
	memset(tx_ring->buffer_info, 0, size);

	memset(tx_ring->desc, 0, tx_ring->size);

	tx_ring->next_to_use = 0;
	tx_ring->next_to_clean = 0;

2294 2295
	writel(0, tx_ring->head);
	writel(0, tx_ring->tail);
2296 2297 2298 2299
}

/**
 * e1000e_free_tx_resources - Free Tx Resources per Queue
2300
 * @tx_ring: Tx descriptor ring
2301 2302 2303
 *
 * Free all transmit software resources
 **/
2304
void e1000e_free_tx_resources(struct e1000_ring *tx_ring)
2305
{
2306
	struct e1000_adapter *adapter = tx_ring->adapter;
2307 2308
	struct pci_dev *pdev = adapter->pdev;

2309
	e1000_clean_tx_ring(tx_ring);
2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320

	vfree(tx_ring->buffer_info);
	tx_ring->buffer_info = NULL;

	dma_free_coherent(&pdev->dev, tx_ring->size, tx_ring->desc,
			  tx_ring->dma);
	tx_ring->desc = NULL;
}

/**
 * e1000e_free_rx_resources - Free Rx Resources
2321
 * @rx_ring: Rx descriptor ring
2322 2323 2324
 *
 * Free all receive software resources
 **/
2325
void e1000e_free_rx_resources(struct e1000_ring *rx_ring)
2326
{
2327
	struct e1000_adapter *adapter = rx_ring->adapter;
2328
	struct pci_dev *pdev = adapter->pdev;
A
Auke Kok 已提交
2329
	int i;
2330

2331
	e1000_clean_rx_ring(rx_ring);
2332

B
Bruce Allan 已提交
2333
	for (i = 0; i < rx_ring->count; i++)
A
Auke Kok 已提交
2334 2335
		kfree(rx_ring->buffer_info[i].ps_pages);

2336 2337 2338 2339 2340 2341 2342 2343 2344 2345
	vfree(rx_ring->buffer_info);
	rx_ring->buffer_info = NULL;

	dma_free_coherent(&pdev->dev, rx_ring->size, rx_ring->desc,
			  rx_ring->dma);
	rx_ring->desc = NULL;
}

/**
 * e1000_update_itr - update the dynamic ITR value based on statistics
2346 2347 2348 2349 2350
 * @adapter: pointer to adapter
 * @itr_setting: current adapter->itr
 * @packets: the number of packets during this measurement interval
 * @bytes: the number of bytes during this measurement interval
 *
2351 2352 2353 2354 2355 2356
 *      Stores a new ITR value based on packets and byte
 *      counts during the last interrupt.  The advantage of per interrupt
 *      computation is faster updates and more accurate ITR for the current
 *      traffic pattern.  Constants in this function were computed
 *      based on theoretical maximum wire speed and thresholds were set based
 *      on testing data as well as attempting to minimize response time
2357 2358
 *      while increasing bulk throughput.  This functionality is controlled
 *      by the InterruptThrottleRate module parameter.
2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373
 **/
static unsigned int e1000_update_itr(struct e1000_adapter *adapter,
				     u16 itr_setting, int packets,
				     int bytes)
{
	unsigned int retval = itr_setting;

	if (packets == 0)
		goto update_itr_done;

	switch (itr_setting) {
	case lowest_latency:
		/* handle TSO and jumbo frames */
		if (bytes/packets > 8000)
			retval = bulk_latency;
B
Bruce Allan 已提交
2374
		else if ((packets < 5) && (bytes > 512))
2375 2376 2377 2378 2379
			retval = low_latency;
		break;
	case low_latency:  /* 50 usec aka 20000 ints/s */
		if (bytes > 10000) {
			/* this if handles the TSO accounting */
B
Bruce Allan 已提交
2380
			if (bytes/packets > 8000)
2381
				retval = bulk_latency;
B
Bruce Allan 已提交
2382
			else if ((packets < 10) || ((bytes/packets) > 1200))
2383
				retval = bulk_latency;
B
Bruce Allan 已提交
2384
			else if ((packets > 35))
2385 2386 2387 2388 2389 2390 2391 2392 2393
				retval = lowest_latency;
		} else if (bytes/packets > 2000) {
			retval = bulk_latency;
		} else if (packets <= 2 && bytes < 512) {
			retval = lowest_latency;
		}
		break;
	case bulk_latency: /* 250 usec aka 4000 ints/s */
		if (bytes > 25000) {
B
Bruce Allan 已提交
2394
			if (packets > 35)
2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418
				retval = low_latency;
		} else if (bytes < 6000) {
			retval = low_latency;
		}
		break;
	}

update_itr_done:
	return retval;
}

static void e1000_set_itr(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u16 current_itr;
	u32 new_itr = adapter->itr;

	/* for non-gigabit speeds, just fix the interrupt rate at 4000 */
	if (adapter->link_speed != SPEED_1000) {
		current_itr = 0;
		new_itr = 4000;
		goto set_itr_now;
	}

2419 2420 2421 2422 2423
	if (adapter->flags2 & FLAG2_DISABLE_AIM) {
		new_itr = 0;
		goto set_itr_now;
	}

2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458
	adapter->tx_itr = e1000_update_itr(adapter,
				    adapter->tx_itr,
				    adapter->total_tx_packets,
				    adapter->total_tx_bytes);
	/* conservative mode (itr 3) eliminates the lowest_latency setting */
	if (adapter->itr_setting == 3 && adapter->tx_itr == lowest_latency)
		adapter->tx_itr = low_latency;

	adapter->rx_itr = e1000_update_itr(adapter,
				    adapter->rx_itr,
				    adapter->total_rx_packets,
				    adapter->total_rx_bytes);
	/* conservative mode (itr 3) eliminates the lowest_latency setting */
	if (adapter->itr_setting == 3 && adapter->rx_itr == lowest_latency)
		adapter->rx_itr = low_latency;

	current_itr = max(adapter->rx_itr, adapter->tx_itr);

	switch (current_itr) {
	/* counts and packets in update_itr are dependent on these numbers */
	case lowest_latency:
		new_itr = 70000;
		break;
	case low_latency:
		new_itr = 20000; /* aka hwitr = ~200 */
		break;
	case bulk_latency:
		new_itr = 4000;
		break;
	default:
		break;
	}

set_itr_now:
	if (new_itr != adapter->itr) {
2459 2460
		/*
		 * this attempts to bias the interrupt rate towards Bulk
2461
		 * by adding intermediate steps when interrupt rate is
2462 2463
		 * increasing
		 */
2464 2465 2466 2467
		new_itr = new_itr > adapter->itr ?
			     min(adapter->itr + (new_itr >> 2), new_itr) :
			     new_itr;
		adapter->itr = new_itr;
2468 2469 2470 2471
		adapter->rx_ring->itr_val = new_itr;
		if (adapter->msix_entries)
			adapter->rx_ring->set_itr = 1;
		else
2472 2473 2474 2475
			if (new_itr)
				ew32(ITR, 1000000000 / (new_itr * 256));
			else
				ew32(ITR, 0);
2476 2477 2478
	}
}

2479 2480 2481 2482 2483 2484
/**
 * e1000_alloc_queues - Allocate memory for all rings
 * @adapter: board private structure to initialize
 **/
static int __devinit e1000_alloc_queues(struct e1000_adapter *adapter)
{
2485 2486 2487
	int size = sizeof(struct e1000_ring);

	adapter->tx_ring = kzalloc(size, GFP_KERNEL);
2488 2489
	if (!adapter->tx_ring)
		goto err;
2490 2491
	adapter->tx_ring->count = adapter->tx_ring_count;
	adapter->tx_ring->adapter = adapter;
2492

2493
	adapter->rx_ring = kzalloc(size, GFP_KERNEL);
2494 2495
	if (!adapter->rx_ring)
		goto err;
2496 2497
	adapter->rx_ring->count = adapter->rx_ring_count;
	adapter->rx_ring->adapter = adapter;
2498 2499 2500 2501 2502 2503 2504 2505 2506

	return 0;
err:
	e_err("Unable to allocate memory for queues\n");
	kfree(adapter->rx_ring);
	kfree(adapter->tx_ring);
	return -ENOMEM;
}

2507 2508
/**
 * e1000_clean - NAPI Rx polling callback
2509
 * @napi: struct associated with this polling callback
2510
 * @budget: amount of packets driver is allowed to process this poll
2511 2512 2513 2514
 **/
static int e1000_clean(struct napi_struct *napi, int budget)
{
	struct e1000_adapter *adapter = container_of(napi, struct e1000_adapter, napi);
2515
	struct e1000_hw *hw = &adapter->hw;
2516
	struct net_device *poll_dev = adapter->netdev;
2517
	int tx_cleaned = 1, work_done = 0;
2518

2519
	adapter = netdev_priv(poll_dev);
2520

2521 2522 2523 2524
	if (adapter->msix_entries &&
	    !(adapter->rx_ring->ims_val & adapter->tx_ring->ims_val))
		goto clean_rx;

2525
	tx_cleaned = e1000_clean_tx_irq(adapter->tx_ring);
2526

2527
clean_rx:
2528
	adapter->clean_rx(adapter->rx_ring, &work_done, budget);
2529

2530
	if (!tx_cleaned)
2531
		work_done = budget;
2532

2533 2534
	/* If budget not fully consumed, exit the polling mode */
	if (work_done < budget) {
2535 2536
		if (adapter->itr_setting & 3)
			e1000_set_itr(adapter);
2537
		napi_complete(napi);
2538 2539 2540 2541 2542 2543
		if (!test_bit(__E1000_DOWN, &adapter->state)) {
			if (adapter->msix_entries)
				ew32(IMS, adapter->rx_ring->ims_val);
			else
				e1000_irq_enable(adapter);
		}
2544 2545 2546 2547 2548
	}

	return work_done;
}

2549
static int e1000_vlan_rx_add_vid(struct net_device *netdev, u16 vid)
2550 2551 2552 2553 2554 2555 2556 2557 2558
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 vfta, index;

	/* don't update vlan cookie if already programmed */
	if ((adapter->hw.mng_cookie.status &
	     E1000_MNG_DHCP_COOKIE_STATUS_VLAN) &&
	    (vid == adapter->mng_vlan_id))
2559
		return 0;
2560

2561
	/* add VID to filter table */
2562 2563 2564 2565 2566 2567
	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER) {
		index = (vid >> 5) & 0x7F;
		vfta = E1000_READ_REG_ARRAY(hw, E1000_VFTA, index);
		vfta |= (1 << (vid & 0x1F));
		hw->mac.ops.write_vfta(hw, index, vfta);
	}
J
Jeff Kirsher 已提交
2568 2569

	set_bit(vid, adapter->active_vlans);
2570 2571

	return 0;
2572 2573
}

2574
static int e1000_vlan_rx_kill_vid(struct net_device *netdev, u16 vid)
2575 2576 2577 2578 2579 2580 2581 2582 2583
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 vfta, index;

	if ((adapter->hw.mng_cookie.status &
	     E1000_MNG_DHCP_COOKIE_STATUS_VLAN) &&
	    (vid == adapter->mng_vlan_id)) {
		/* release control to f/w */
2584
		e1000e_release_hw_control(adapter);
2585
		return 0;
2586 2587 2588
	}

	/* remove VID from filter table */
2589 2590 2591 2592 2593 2594
	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER) {
		index = (vid >> 5) & 0x7F;
		vfta = E1000_READ_REG_ARRAY(hw, E1000_VFTA, index);
		vfta &= ~(1 << (vid & 0x1F));
		hw->mac.ops.write_vfta(hw, index, vfta);
	}
J
Jeff Kirsher 已提交
2595 2596

	clear_bit(vid, adapter->active_vlans);
2597 2598

	return 0;
2599 2600
}

J
Jeff Kirsher 已提交
2601 2602 2603 2604 2605
/**
 * e1000e_vlan_filter_disable - helper to disable hw VLAN filtering
 * @adapter: board private structure to initialize
 **/
static void e1000e_vlan_filter_disable(struct e1000_adapter *adapter)
2606 2607
{
	struct net_device *netdev = adapter->netdev;
J
Jeff Kirsher 已提交
2608 2609
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl;
2610

J
Jeff Kirsher 已提交
2611 2612 2613 2614 2615 2616 2617 2618 2619
	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER) {
		/* disable VLAN receive filtering */
		rctl = er32(RCTL);
		rctl &= ~(E1000_RCTL_VFE | E1000_RCTL_CFIEN);
		ew32(RCTL, rctl);

		if (adapter->mng_vlan_id != (u16)E1000_MNG_VLAN_NONE) {
			e1000_vlan_rx_kill_vid(netdev, adapter->mng_vlan_id);
			adapter->mng_vlan_id = E1000_MNG_VLAN_NONE;
2620 2621 2622 2623
		}
	}
}

J
Jeff Kirsher 已提交
2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640
/**
 * e1000e_vlan_filter_enable - helper to enable HW VLAN filtering
 * @adapter: board private structure to initialize
 **/
static void e1000e_vlan_filter_enable(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl;

	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER) {
		/* enable VLAN receive filtering */
		rctl = er32(RCTL);
		rctl |= E1000_RCTL_VFE;
		rctl &= ~E1000_RCTL_CFIEN;
		ew32(RCTL, rctl);
	}
}
2641

J
Jeff Kirsher 已提交
2642 2643 2644 2645 2646
/**
 * e1000e_vlan_strip_enable - helper to disable HW VLAN stripping
 * @adapter: board private structure to initialize
 **/
static void e1000e_vlan_strip_disable(struct e1000_adapter *adapter)
2647 2648
{
	struct e1000_hw *hw = &adapter->hw;
J
Jeff Kirsher 已提交
2649
	u32 ctrl;
2650

J
Jeff Kirsher 已提交
2651 2652 2653 2654 2655
	/* disable VLAN tag insert/strip */
	ctrl = er32(CTRL);
	ctrl &= ~E1000_CTRL_VME;
	ew32(CTRL, ctrl);
}
2656

J
Jeff Kirsher 已提交
2657 2658 2659 2660 2661 2662 2663 2664
/**
 * e1000e_vlan_strip_enable - helper to enable HW VLAN stripping
 * @adapter: board private structure to initialize
 **/
static void e1000e_vlan_strip_enable(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl;
2665

J
Jeff Kirsher 已提交
2666 2667 2668 2669 2670
	/* enable VLAN tag insert/strip */
	ctrl = er32(CTRL);
	ctrl |= E1000_CTRL_VME;
	ew32(CTRL, ctrl);
}
2671

J
Jeff Kirsher 已提交
2672 2673 2674 2675 2676 2677 2678 2679 2680 2681
static void e1000_update_mng_vlan(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	u16 vid = adapter->hw.mng_cookie.vlan_id;
	u16 old_vid = adapter->mng_vlan_id;

	if (adapter->hw.mng_cookie.status &
	    E1000_MNG_DHCP_COOKIE_STATUS_VLAN) {
		e1000_vlan_rx_add_vid(netdev, vid);
		adapter->mng_vlan_id = vid;
2682 2683
	}

J
Jeff Kirsher 已提交
2684 2685
	if ((old_vid != (u16)E1000_MNG_VLAN_NONE) && (vid != old_vid))
		e1000_vlan_rx_kill_vid(netdev, old_vid);
2686 2687 2688 2689 2690 2691
}

static void e1000_restore_vlan(struct e1000_adapter *adapter)
{
	u16 vid;

J
Jeff Kirsher 已提交
2692
	e1000_vlan_rx_add_vid(adapter->netdev, 0);
2693

J
Jeff Kirsher 已提交
2694
	for_each_set_bit(vid, adapter->active_vlans, VLAN_N_VID)
2695 2696 2697
		e1000_vlan_rx_add_vid(adapter->netdev, vid);
}

2698
static void e1000_init_manageability_pt(struct e1000_adapter *adapter)
2699 2700
{
	struct e1000_hw *hw = &adapter->hw;
2701
	u32 manc, manc2h, mdef, i, j;
2702 2703 2704 2705 2706 2707

	if (!(adapter->flags & FLAG_MNG_PT_ENABLED))
		return;

	manc = er32(MANC);

2708 2709
	/*
	 * enable receiving management packets to the host. this will probably
2710
	 * generate destination unreachable messages from the host OS, but
2711 2712
	 * the packets will be handled on SMBUS
	 */
2713 2714
	manc |= E1000_MANC_EN_MNG2HOST;
	manc2h = er32(MANC2H);
2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729

	switch (hw->mac.type) {
	default:
		manc2h |= (E1000_MANC2H_PORT_623 | E1000_MANC2H_PORT_664);
		break;
	case e1000_82574:
	case e1000_82583:
		/*
		 * Check if IPMI pass-through decision filter already exists;
		 * if so, enable it.
		 */
		for (i = 0, j = 0; i < 8; i++) {
			mdef = er32(MDEF(i));

			/* Ignore filters with anything other than IPMI ports */
2730
			if (mdef & ~(E1000_MDEF_PORT_623 | E1000_MDEF_PORT_664))
2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757
				continue;

			/* Enable this decision filter in MANC2H */
			if (mdef)
				manc2h |= (1 << i);

			j |= mdef;
		}

		if (j == (E1000_MDEF_PORT_623 | E1000_MDEF_PORT_664))
			break;

		/* Create new decision filter in an empty filter */
		for (i = 0, j = 0; i < 8; i++)
			if (er32(MDEF(i)) == 0) {
				ew32(MDEF(i), (E1000_MDEF_PORT_623 |
					       E1000_MDEF_PORT_664));
				manc2h |= (1 << 1);
				j++;
				break;
			}

		if (!j)
			e_warn("Unable to create IPMI pass-through filter\n");
		break;
	}

2758 2759 2760 2761 2762
	ew32(MANC2H, manc2h);
	ew32(MANC, manc);
}

/**
2763
 * e1000_configure_tx - Configure Transmit Unit after Reset
2764 2765 2766 2767 2768 2769 2770 2771 2772
 * @adapter: board private structure
 *
 * Configure the Tx unit of the MAC after a reset.
 **/
static void e1000_configure_tx(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_ring *tx_ring = adapter->tx_ring;
	u64 tdba;
2773
	u32 tdlen, tarc;
2774 2775 2776 2777

	/* Setup the HW Tx Head and Tail descriptor pointers */
	tdba = tx_ring->dma;
	tdlen = tx_ring->count * sizeof(struct e1000_tx_desc);
2778
	ew32(TDBAL, (tdba & DMA_BIT_MASK(32)));
2779 2780 2781 2782
	ew32(TDBAH, (tdba >> 32));
	ew32(TDLEN, tdlen);
	ew32(TDH, 0);
	ew32(TDT, 0);
2783 2784
	tx_ring->head = adapter->hw.hw_addr + E1000_TDH;
	tx_ring->tail = adapter->hw.hw_addr + E1000_TDT;
2785 2786 2787

	/* Set the Tx Interrupt Delay register */
	ew32(TIDV, adapter->tx_int_delay);
2788
	/* Tx irq moderation */
2789 2790
	ew32(TADV, adapter->tx_abs_int_delay);

2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802
	if (adapter->flags2 & FLAG2_DMA_BURST) {
		u32 txdctl = er32(TXDCTL(0));
		txdctl &= ~(E1000_TXDCTL_PTHRESH | E1000_TXDCTL_HTHRESH |
			    E1000_TXDCTL_WTHRESH);
		/*
		 * set up some performance related parameters to encourage the
		 * hardware to use the bus more efficiently in bursts, depends
		 * on the tx_int_delay to be enabled,
		 * wthresh = 5 ==> burst write a cacheline (64 bytes) at a time
		 * hthresh = 1 ==> prefetch when one or more available
		 * pthresh = 0x1f ==> prefetch if internal cache 31 or less
		 * BEWARE: this seems to work but should be considered first if
2803
		 * there are Tx hangs or other Tx related bugs
2804 2805 2806 2807
		 */
		txdctl |= E1000_TXDCTL_DMA_BURST_ENABLE;
		ew32(TXDCTL(0), txdctl);
	}
2808 2809
	/* erratum work around: set txdctl the same for both queues */
	ew32(TXDCTL(1), er32(TXDCTL(0)));
2810

2811
	if (adapter->flags & FLAG_TARC_SPEED_MODE_BIT) {
2812
		tarc = er32(TARC(0));
2813 2814 2815 2816
		/*
		 * set the speed mode bit, we'll clear it if we're not at
		 * gigabit link later
		 */
2817 2818
#define SPEED_MODE_BIT (1 << 21)
		tarc |= SPEED_MODE_BIT;
2819
		ew32(TARC(0), tarc);
2820 2821 2822 2823
	}

	/* errata: program both queues to unweighted RR */
	if (adapter->flags & FLAG_TARC_SET_BIT_ZERO) {
2824
		tarc = er32(TARC(0));
2825
		tarc |= 1;
2826 2827
		ew32(TARC(0), tarc);
		tarc = er32(TARC(1));
2828
		tarc |= 1;
2829
		ew32(TARC(1), tarc);
2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841
	}

	/* Setup Transmit Descriptor Settings for eop descriptor */
	adapter->txd_cmd = E1000_TXD_CMD_EOP | E1000_TXD_CMD_IFCS;

	/* only set IDE if we are delaying interrupts using the timers */
	if (adapter->tx_int_delay)
		adapter->txd_cmd |= E1000_TXD_CMD_IDE;

	/* enable Report Status bit */
	adapter->txd_cmd |= E1000_TXD_CMD_RS;

2842
	e1000e_config_collision_dist(hw);
2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856
}

/**
 * e1000_setup_rctl - configure the receive control registers
 * @adapter: Board private structure
 **/
#define PAGE_USE_COUNT(S) (((S) >> PAGE_SHIFT) + \
			   (((S) & (PAGE_SIZE - 1)) ? 1 : 0))
static void e1000_setup_rctl(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl, rfctl;
	u32 pages = 0;

2857 2858 2859 2860 2861 2862 2863 2864
	/* Workaround Si errata on 82579 - configure jumbo frame flow */
	if (hw->mac.type == e1000_pch2lan) {
		s32 ret_val;

		if (adapter->netdev->mtu > ETH_DATA_LEN)
			ret_val = e1000_lv_jumbo_workaround_ich8lan(hw, true);
		else
			ret_val = e1000_lv_jumbo_workaround_ich8lan(hw, false);
2865 2866 2867

		if (ret_val)
			e_dbg("failed to enable jumbo frame workaround mode\n");
2868 2869
	}

2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885
	/* Program MC offset vector base */
	rctl = er32(RCTL);
	rctl &= ~(3 << E1000_RCTL_MO_SHIFT);
	rctl |= E1000_RCTL_EN | E1000_RCTL_BAM |
		E1000_RCTL_LBM_NO | E1000_RCTL_RDMTS_HALF |
		(adapter->hw.mac.mc_filter_type << E1000_RCTL_MO_SHIFT);

	/* Do not Store bad packets */
	rctl &= ~E1000_RCTL_SBP;

	/* Enable Long Packet receive */
	if (adapter->netdev->mtu <= ETH_DATA_LEN)
		rctl &= ~E1000_RCTL_LPE;
	else
		rctl |= E1000_RCTL_LPE;

J
Jeff Kirsher 已提交
2886 2887 2888 2889 2890 2891
	/* Some systems expect that the CRC is included in SMBUS traffic. The
	 * hardware strips the CRC before sending to both SMBUS (BMC) and to
	 * host memory when this is enabled
	 */
	if (adapter->flags2 & FLAG2_CRC_STRIPPING)
		rctl |= E1000_RCTL_SECRC;
2892

2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909
	/* Workaround Si errata on 82577 PHY - configure IPG for jumbos */
	if ((hw->phy.type == e1000_phy_82577) && (rctl & E1000_RCTL_LPE)) {
		u16 phy_data;

		e1e_rphy(hw, PHY_REG(770, 26), &phy_data);
		phy_data &= 0xfff8;
		phy_data |= (1 << 2);
		e1e_wphy(hw, PHY_REG(770, 26), phy_data);

		e1e_rphy(hw, 22, &phy_data);
		phy_data &= 0x0fff;
		phy_data |= (1 << 14);
		e1e_wphy(hw, 0x10, 0x2823);
		e1e_wphy(hw, 0x11, 0x0003);
		e1e_wphy(hw, 22, phy_data);
	}

2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929
	/* Setup buffer sizes */
	rctl &= ~E1000_RCTL_SZ_4096;
	rctl |= E1000_RCTL_BSEX;
	switch (adapter->rx_buffer_len) {
	case 2048:
	default:
		rctl |= E1000_RCTL_SZ_2048;
		rctl &= ~E1000_RCTL_BSEX;
		break;
	case 4096:
		rctl |= E1000_RCTL_SZ_4096;
		break;
	case 8192:
		rctl |= E1000_RCTL_SZ_8192;
		break;
	case 16384:
		rctl |= E1000_RCTL_SZ_16384;
		break;
	}

2930 2931 2932 2933
	/* Enable Extended Status in all Receive Descriptors */
	rfctl = er32(RFCTL);
	rfctl |= E1000_RFCTL_EXTEN;

2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949
	/*
	 * 82571 and greater support packet-split where the protocol
	 * header is placed in skb->data and the packet data is
	 * placed in pages hanging off of skb_shinfo(skb)->nr_frags.
	 * In the case of a non-split, skb->data is linearly filled,
	 * followed by the page buffers.  Therefore, skb->data is
	 * sized to hold the largest protocol header.
	 *
	 * allocations using alloc_page take too long for regular MTU
	 * so only enable packet split for jumbo frames
	 *
	 * Using pages when the page size is greater than 16k wastes
	 * a lot of memory, since we allocate 3 pages at all times
	 * per packet.
	 */
	pages = PAGE_USE_COUNT(adapter->netdev->mtu);
2950
	if ((pages <= 3) && (PAGE_SIZE <= 16384) && (rctl & E1000_RCTL_LPE))
2951
		adapter->rx_ps_pages = pages;
2952 2953
	else
		adapter->rx_ps_pages = 0;
2954 2955

	if (adapter->rx_ps_pages) {
2956 2957
		u32 psrctl = 0;

2958 2959 2960 2961
		/*
		 * disable packet split support for IPv6 extension headers,
		 * because some malformed IPv6 headers can hang the Rx
		 */
2962 2963 2964
		rfctl |= (E1000_RFCTL_IPV6_EX_DIS |
			  E1000_RFCTL_NEW_IPV6_EXT_DIS);

A
Auke Kok 已提交
2965 2966
		/* Enable Packet split descriptors */
		rctl |= E1000_RCTL_DTYP_PS;
2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986

		psrctl |= adapter->rx_ps_bsize0 >>
			E1000_PSRCTL_BSIZE0_SHIFT;

		switch (adapter->rx_ps_pages) {
		case 3:
			psrctl |= PAGE_SIZE <<
				E1000_PSRCTL_BSIZE3_SHIFT;
		case 2:
			psrctl |= PAGE_SIZE <<
				E1000_PSRCTL_BSIZE2_SHIFT;
		case 1:
			psrctl |= PAGE_SIZE >>
				E1000_PSRCTL_BSIZE1_SHIFT;
			break;
		}

		ew32(PSRCTL, psrctl);
	}

2987
	ew32(RFCTL, rfctl);
2988
	ew32(RCTL, rctl);
2989 2990
	/* just started the receive unit, no need to restart */
	adapter->flags &= ~FLAG_RX_RESTART_NOW;
2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008
}

/**
 * e1000_configure_rx - Configure Receive Unit after Reset
 * @adapter: board private structure
 *
 * Configure the Rx unit of the MAC after a reset.
 **/
static void e1000_configure_rx(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_ring *rx_ring = adapter->rx_ring;
	u64 rdba;
	u32 rdlen, rctl, rxcsum, ctrl_ext;

	if (adapter->rx_ps_pages) {
		/* this is a 32 byte descriptor */
		rdlen = rx_ring->count *
3009
		    sizeof(union e1000_rx_desc_packet_split);
3010 3011
		adapter->clean_rx = e1000_clean_rx_irq_ps;
		adapter->alloc_rx_buf = e1000_alloc_rx_buffers_ps;
3012
	} else if (adapter->netdev->mtu > ETH_FRAME_LEN + ETH_FCS_LEN) {
3013
		rdlen = rx_ring->count * sizeof(union e1000_rx_desc_extended);
3014 3015
		adapter->clean_rx = e1000_clean_jumbo_rx_irq;
		adapter->alloc_rx_buf = e1000_alloc_jumbo_rx_buffers;
3016
	} else {
3017
		rdlen = rx_ring->count * sizeof(union e1000_rx_desc_extended);
3018 3019 3020 3021 3022 3023
		adapter->clean_rx = e1000_clean_rx_irq;
		adapter->alloc_rx_buf = e1000_alloc_rx_buffers;
	}

	/* disable receives while setting up the descriptors */
	rctl = er32(RCTL);
3024 3025
	if (!(adapter->flags2 & FLAG2_NO_DISABLE_RX))
		ew32(RCTL, rctl & ~E1000_RCTL_EN);
3026
	e1e_flush();
3027
	usleep_range(10000, 20000);
3028

3029 3030 3031 3032
	if (adapter->flags2 & FLAG2_DMA_BURST) {
		/*
		 * set the writeback threshold (only takes effect if the RDTR
		 * is set). set GRAN=1 and write back up to 0x4 worth, and
3033
		 * enable prefetching of 0x20 Rx descriptors
3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051
		 * granularity = 01
		 * wthresh = 04,
		 * hthresh = 04,
		 * pthresh = 0x20
		 */
		ew32(RXDCTL(0), E1000_RXDCTL_DMA_BURST_ENABLE);
		ew32(RXDCTL(1), E1000_RXDCTL_DMA_BURST_ENABLE);

		/*
		 * override the delay timers for enabling bursting, only if
		 * the value was not set by the user via module options
		 */
		if (adapter->rx_int_delay == DEFAULT_RDTR)
			adapter->rx_int_delay = BURST_RDTR;
		if (adapter->rx_abs_int_delay == DEFAULT_RADV)
			adapter->rx_abs_int_delay = BURST_RADV;
	}

3052 3053 3054 3055 3056
	/* set the Receive Delay Timer Register */
	ew32(RDTR, adapter->rx_int_delay);

	/* irq moderation */
	ew32(RADV, adapter->rx_abs_int_delay);
3057
	if ((adapter->itr_setting != 0) && (adapter->itr != 0))
3058
		ew32(ITR, 1000000000 / (adapter->itr * 256));
3059 3060 3061 3062 3063 3064 3065 3066

	ctrl_ext = er32(CTRL_EXT);
	/* Auto-Mask interrupts upon ICR access */
	ctrl_ext |= E1000_CTRL_EXT_IAME;
	ew32(IAM, 0xffffffff);
	ew32(CTRL_EXT, ctrl_ext);
	e1e_flush();

3067 3068 3069 3070
	/*
	 * Setup the HW Rx Head and Tail Descriptor Pointers and
	 * the Base and Length of the Rx Descriptor Ring
	 */
3071
	rdba = rx_ring->dma;
3072
	ew32(RDBAL, (rdba & DMA_BIT_MASK(32)));
3073 3074 3075 3076
	ew32(RDBAH, (rdba >> 32));
	ew32(RDLEN, rdlen);
	ew32(RDH, 0);
	ew32(RDT, 0);
3077 3078
	rx_ring->head = adapter->hw.hw_addr + E1000_RDH;
	rx_ring->tail = adapter->hw.hw_addr + E1000_RDT;
3079 3080 3081

	/* Enable Receive Checksum Offload for TCP and UDP */
	rxcsum = er32(RXCSUM);
3082
	if (adapter->netdev->features & NETIF_F_RXCSUM) {
3083 3084
		rxcsum |= E1000_RXCSUM_TUOFL;

3085 3086 3087 3088
		/*
		 * IPv4 payload checksum for UDP fragments must be
		 * used in conjunction with packet-split.
		 */
3089 3090 3091 3092 3093 3094 3095 3096
		if (adapter->rx_ps_pages)
			rxcsum |= E1000_RXCSUM_IPPCSE;
	} else {
		rxcsum &= ~E1000_RXCSUM_TUOFL;
		/* no need to clear IPPCSE as it defaults to 0 */
	}
	ew32(RXCSUM, rxcsum);

3097 3098 3099 3100 3101
	if (adapter->hw.mac.type == e1000_pch2lan) {
		/*
		 * With jumbo frames, excessive C-state transition
		 * latencies result in dropped transactions.
		 */
3102 3103 3104
		if (adapter->netdev->mtu > ETH_DATA_LEN) {
			u32 rxdctl = er32(RXDCTL(0));
			ew32(RXDCTL(0), rxdctl | 0x3);
3105
			pm_qos_update_request(&adapter->netdev->pm_qos_req, 55);
3106
		} else {
3107 3108
			pm_qos_update_request(&adapter->netdev->pm_qos_req,
					      PM_QOS_DEFAULT_VALUE);
3109
		}
3110
	}
3111 3112 3113 3114 3115 3116

	/* Enable Receives */
	ew32(RCTL, rctl);
}

/**
3117 3118
 * e1000e_write_mc_addr_list - write multicast addresses to MTA
 * @netdev: network interface device structure
3119
 *
3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156
 * Writes multicast address list to the MTA hash table.
 * Returns: -ENOMEM on failure
 *                0 on no addresses written
 *                X on writing X addresses to MTA
 */
static int e1000e_write_mc_addr_list(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	struct netdev_hw_addr *ha;
	u8 *mta_list;
	int i;

	if (netdev_mc_empty(netdev)) {
		/* nothing to program, so clear mc list */
		hw->mac.ops.update_mc_addr_list(hw, NULL, 0);
		return 0;
	}

	mta_list = kzalloc(netdev_mc_count(netdev) * ETH_ALEN, GFP_ATOMIC);
	if (!mta_list)
		return -ENOMEM;

	/* update_mc_addr_list expects a packed array of only addresses. */
	i = 0;
	netdev_for_each_mc_addr(ha, netdev)
		memcpy(mta_list + (i++ * ETH_ALEN), ha->addr, ETH_ALEN);

	hw->mac.ops.update_mc_addr_list(hw, mta_list, i);
	kfree(mta_list);

	return netdev_mc_count(netdev);
}

/**
 * e1000e_write_uc_addr_list - write unicast addresses to RAR table
 * @netdev: network interface device structure
3157
 *
3158 3159 3160 3161
 * Writes unicast address list to the RAR table.
 * Returns: -ENOMEM on failure/insufficient address space
 *                0 on no addresses written
 *                X on writing X addresses to the RAR table
3162
 **/
3163
static int e1000e_write_uc_addr_list(struct net_device *netdev)
3164
{
3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	unsigned int rar_entries = hw->mac.rar_entry_count;
	int count = 0;

	/* save a rar entry for our hardware address */
	rar_entries--;

	/* save a rar entry for the LAA workaround */
	if (adapter->flags & FLAG_RESET_OVERWRITES_LAA)
		rar_entries--;

	/* return ENOMEM indicating insufficient memory for addresses */
	if (netdev_uc_count(netdev) > rar_entries)
		return -ENOMEM;

	if (!netdev_uc_empty(netdev) && rar_entries) {
		struct netdev_hw_addr *ha;

		/*
		 * write the addresses in reverse order to avoid write
		 * combining
		 */
		netdev_for_each_uc_addr(ha, netdev) {
			if (!rar_entries)
				break;
			e1000e_rar_set(hw, ha->addr, rar_entries--);
			count++;
		}
	}

	/* zero out the remaining RAR entries not used above */
	for (; rar_entries > 0; rar_entries--) {
		ew32(RAH(rar_entries), 0);
		ew32(RAL(rar_entries), 0);
	}
	e1e_flush();

	return count;
3204 3205 3206
}

/**
3207
 * e1000e_set_rx_mode - secondary unicast, Multicast and Promiscuous mode set
3208 3209
 * @netdev: network interface device structure
 *
3210 3211 3212
 * The ndo_set_rx_mode entry point is called whenever the unicast or multicast
 * address list or the network interface flags are updated.  This routine is
 * responsible for configuring the hardware for proper unicast, multicast,
3213 3214
 * promiscuous mode, and all-multi behavior.
 **/
3215
static void e1000e_set_rx_mode(struct net_device *netdev)
3216 3217 3218 3219 3220 3221 3222 3223
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl;

	/* Check for Promiscuous and All Multicast modes */
	rctl = er32(RCTL);

3224 3225 3226
	/* clear the affected bits */
	rctl &= ~(E1000_RCTL_UPE | E1000_RCTL_MPE);

3227 3228
	if (netdev->flags & IFF_PROMISC) {
		rctl |= (E1000_RCTL_UPE | E1000_RCTL_MPE);
J
Jeff Kirsher 已提交
3229 3230
		/* Do not hardware filter VLANs in promisc mode */
		e1000e_vlan_filter_disable(adapter);
3231
	} else {
3232
		int count;
3233 3234 3235
		if (netdev->flags & IFF_ALLMULTI) {
			rctl |= E1000_RCTL_MPE;
		} else {
3236 3237 3238 3239 3240 3241 3242 3243
			/*
			 * Write addresses to the MTA, if the attempt fails
			 * then we should just turn on promiscuous mode so
			 * that we can at least receive multicast traffic
			 */
			count = e1000e_write_mc_addr_list(netdev);
			if (count < 0)
				rctl |= E1000_RCTL_MPE;
3244
		}
J
Jeff Kirsher 已提交
3245
		e1000e_vlan_filter_enable(adapter);
3246
		/*
3247 3248 3249
		 * Write addresses to available RAR registers, if there is not
		 * sufficient space to store all the addresses then enable
		 * unicast promiscuous mode
3250
		 */
3251 3252 3253
		count = e1000e_write_uc_addr_list(netdev);
		if (count < 0)
			rctl |= E1000_RCTL_UPE;
3254
	}
J
Jeff Kirsher 已提交
3255

3256 3257
	ew32(RCTL, rctl);

J
Jeff Kirsher 已提交
3258 3259 3260 3261
	if (netdev->features & NETIF_F_HW_VLAN_RX)
		e1000e_vlan_strip_enable(adapter);
	else
		e1000e_vlan_strip_disable(adapter);
3262 3263
}

3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299
static void e1000e_setup_rss_hash(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 mrqc, rxcsum;
	int i;
	static const u32 rsskey[10] = {
		0xda565a6d, 0xc20e5b25, 0x3d256741, 0xb08fa343, 0xcb2bcad0,
		0xb4307bae, 0xa32dcb77, 0x0cf23080, 0x3bb7426a, 0xfa01acbe
	};

	/* Fill out hash function seed */
	for (i = 0; i < 10; i++)
		ew32(RSSRK(i), rsskey[i]);

	/* Direct all traffic to queue 0 */
	for (i = 0; i < 32; i++)
		ew32(RETA(i), 0);

	/*
	 * Disable raw packet checksumming so that RSS hash is placed in
	 * descriptor on writeback.
	 */
	rxcsum = er32(RXCSUM);
	rxcsum |= E1000_RXCSUM_PCSD;

	ew32(RXCSUM, rxcsum);

	mrqc = (E1000_MRQC_RSS_FIELD_IPV4 |
		E1000_MRQC_RSS_FIELD_IPV4_TCP |
		E1000_MRQC_RSS_FIELD_IPV6 |
		E1000_MRQC_RSS_FIELD_IPV6_TCP |
		E1000_MRQC_RSS_FIELD_IPV6_TCP_EX);

	ew32(MRQC, mrqc);
}

3300
/**
3301
 * e1000_configure - configure the hardware for Rx and Tx
3302 3303 3304 3305
 * @adapter: private board structure
 **/
static void e1000_configure(struct e1000_adapter *adapter)
{
3306 3307
	struct e1000_ring *rx_ring = adapter->rx_ring;

3308
	e1000e_set_rx_mode(adapter->netdev);
3309 3310

	e1000_restore_vlan(adapter);
3311
	e1000_init_manageability_pt(adapter);
3312 3313

	e1000_configure_tx(adapter);
3314 3315 3316

	if (adapter->netdev->features & NETIF_F_RXHASH)
		e1000e_setup_rss_hash(adapter);
3317 3318
	e1000_setup_rctl(adapter);
	e1000_configure_rx(adapter);
3319
	adapter->alloc_rx_buf(rx_ring, e1000_desc_unused(rx_ring), GFP_KERNEL);
3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331
}

/**
 * e1000e_power_up_phy - restore link in case the phy was powered down
 * @adapter: address of board private structure
 *
 * The phy may be powered down to save power and turn off link when the
 * driver is unloaded and wake on lan is not enabled (among others)
 * *** this routine MUST be followed by a call to e1000e_reset ***
 **/
void e1000e_power_up_phy(struct e1000_adapter *adapter)
{
3332 3333
	if (adapter->hw.phy.ops.power_up)
		adapter->hw.phy.ops.power_up(&adapter->hw);
3334 3335 3336 3337 3338 3339 3340

	adapter->hw.mac.ops.setup_link(&adapter->hw);
}

/**
 * e1000_power_down_phy - Power down the PHY
 *
3341 3342
 * Power down the PHY so no link is implied when interface is down.
 * The PHY cannot be powered down if management or WoL is active.
3343 3344 3345 3346
 */
static void e1000_power_down_phy(struct e1000_adapter *adapter)
{
	/* WoL is enabled */
3347
	if (adapter->wol)
3348 3349
		return;

3350 3351
	if (adapter->hw.phy.ops.power_down)
		adapter->hw.phy.ops.power_down(&adapter->hw);
3352 3353 3354 3355 3356 3357 3358 3359
}

/**
 * e1000e_reset - bring the hardware into a known good state
 *
 * This function boots the hardware and enables some settings that
 * require a configuration cycle of the hardware - those cannot be
 * set/changed during runtime. After reset the device needs to be
3360
 * properly configured for Rx, Tx etc.
3361 3362 3363 3364
 */
void e1000e_reset(struct e1000_adapter *adapter)
{
	struct e1000_mac_info *mac = &adapter->hw.mac;
3365
	struct e1000_fc_info *fc = &adapter->hw.fc;
3366 3367
	struct e1000_hw *hw = &adapter->hw;
	u32 tx_space, min_tx_space, min_rx_space;
3368
	u32 pba = adapter->pba;
3369 3370
	u16 hwm;

3371
	/* reset Packet Buffer Allocation to default */
3372
	ew32(PBA, pba);
3373

3374
	if (adapter->max_frame_size > ETH_FRAME_LEN + ETH_FCS_LEN) {
3375 3376
		/*
		 * To maintain wire speed transmits, the Tx FIFO should be
3377 3378 3379 3380
		 * large enough to accommodate two full transmit packets,
		 * rounded up to the next 1KB and expressed in KB.  Likewise,
		 * the Rx FIFO should be large enough to accommodate at least
		 * one full receive packet and is similarly rounded up and
3381 3382
		 * expressed in KB.
		 */
3383
		pba = er32(PBA);
3384
		/* upper 16 bits has Tx packet buffer allocation size in KB */
3385
		tx_space = pba >> 16;
3386
		/* lower 16 bits has Rx packet buffer allocation size in KB */
3387
		pba &= 0xffff;
3388
		/*
3389
		 * the Tx fifo also stores 16 bytes of information about the Tx
3390
		 * but don't include ethernet FCS because hardware appends it
3391 3392
		 */
		min_tx_space = (adapter->max_frame_size +
3393 3394 3395 3396 3397
				sizeof(struct e1000_tx_desc) -
				ETH_FCS_LEN) * 2;
		min_tx_space = ALIGN(min_tx_space, 1024);
		min_tx_space >>= 10;
		/* software strips receive CRC, so leave room for it */
3398
		min_rx_space = adapter->max_frame_size;
3399 3400 3401
		min_rx_space = ALIGN(min_rx_space, 1024);
		min_rx_space >>= 10;

3402 3403
		/*
		 * If current Tx allocation is less than the min Tx FIFO size,
3404
		 * and the min Tx FIFO size is less than the current Rx FIFO
3405 3406
		 * allocation, take space away from current Rx allocation
		 */
3407 3408 3409
		if ((tx_space < min_tx_space) &&
		    ((min_tx_space - tx_space) < pba)) {
			pba -= min_tx_space - tx_space;
3410

3411
			/*
3412
			 * if short on Rx space, Rx wins and must trump Tx
3413 3414
			 * adjustment or use Early Receive if available
			 */
3415
			if (pba < min_rx_space)
3416
				pba = min_rx_space;
3417
		}
3418 3419

		ew32(PBA, pba);
3420 3421
	}

3422 3423 3424
	/*
	 * flow control settings
	 *
3425
	 * The high water mark must be low enough to fit one full frame
3426 3427 3428
	 * (or the size used for early receive) above it in the Rx FIFO.
	 * Set it to the lower of:
	 * - 90% of the Rx FIFO size, and
3429
	 * - the full Rx FIFO size minus one full frame
3430
	 */
3431 3432 3433 3434 3435 3436 3437 3438
	if (adapter->flags & FLAG_DISABLE_FC_PAUSE_TIME)
		fc->pause_time = 0xFFFF;
	else
		fc->pause_time = E1000_FC_PAUSE_TIME;
	fc->send_xon = 1;
	fc->current_mode = fc->requested_mode;

	switch (hw->mac.type) {
3439 3440 3441 3442 3443 3444 3445 3446 3447 3448
	case e1000_ich9lan:
	case e1000_ich10lan:
		if (adapter->netdev->mtu > ETH_DATA_LEN) {
			pba = 14;
			ew32(PBA, pba);
			fc->high_water = 0x2800;
			fc->low_water = fc->high_water - 8;
			break;
		}
		/* fall-through */
3449
	default:
3450 3451
		hwm = min(((pba << 10) * 9 / 10),
			  ((pba << 10) - adapter->max_frame_size));
3452 3453 3454 3455 3456

		fc->high_water = hwm & E1000_FCRTH_RTH; /* 8-byte granularity */
		fc->low_water = fc->high_water - 8;
		break;
	case e1000_pchlan:
3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467
		/*
		 * Workaround PCH LOM adapter hangs with certain network
		 * loads.  If hangs persist, try disabling Tx flow control.
		 */
		if (adapter->netdev->mtu > ETH_DATA_LEN) {
			fc->high_water = 0x3500;
			fc->low_water  = 0x1500;
		} else {
			fc->high_water = 0x5000;
			fc->low_water  = 0x3000;
		}
3468
		fc->refresh_time = 0x1000;
3469 3470 3471 3472 3473 3474
		break;
	case e1000_pch2lan:
		fc->high_water = 0x05C20;
		fc->low_water = 0x05048;
		fc->pause_time = 0x0650;
		fc->refresh_time = 0x0400;
3475 3476 3477 3478
		if (adapter->netdev->mtu > ETH_DATA_LEN) {
			pba = 14;
			ew32(PBA, pba);
		}
3479
		break;
3480
	}
3481

3482 3483
	/*
	 * Disable Adaptive Interrupt Moderation if 2 full packets cannot
3484
	 * fit in receive buffer.
3485 3486
	 */
	if (adapter->itr_setting & 0x3) {
3487
		if ((adapter->max_frame_size * 2) > (pba << 10)) {
3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502
			if (!(adapter->flags2 & FLAG2_DISABLE_AIM)) {
				dev_info(&adapter->pdev->dev,
					"Interrupt Throttle Rate turned off\n");
				adapter->flags2 |= FLAG2_DISABLE_AIM;
				ew32(ITR, 0);
			}
		} else if (adapter->flags2 & FLAG2_DISABLE_AIM) {
			dev_info(&adapter->pdev->dev,
				 "Interrupt Throttle Rate turned on\n");
			adapter->flags2 &= ~FLAG2_DISABLE_AIM;
			adapter->itr = 20000;
			ew32(ITR, 1000000000 / (adapter->itr * 256));
		}
	}

3503 3504
	/* Allow time for pending master requests to run */
	mac->ops.reset_hw(hw);
3505 3506 3507 3508 3509

	/*
	 * For parts with AMT enabled, let the firmware know
	 * that the network interface is in control
	 */
J
Jesse Brandeburg 已提交
3510
	if (adapter->flags & FLAG_HAS_AMT)
3511
		e1000e_get_hw_control(adapter);
3512

3513 3514 3515
	ew32(WUC, 0);

	if (mac->ops.init_hw(hw))
3516
		e_err("Hardware Error\n");
3517 3518 3519 3520 3521 3522 3523

	e1000_update_mng_vlan(adapter);

	/* Enable h/w to recognize an 802.1Q VLAN Ethernet packet */
	ew32(VET, ETH_P_8021Q);

	e1000e_reset_adaptive(hw);
3524 3525 3526 3527 3528 3529 3530

	if (!netif_running(adapter->netdev) &&
	    !test_bit(__E1000_TESTING, &adapter->state)) {
		e1000_power_down_phy(adapter);
		return;
	}

3531 3532
	e1000_get_phy_info(hw);

3533 3534
	if ((adapter->flags & FLAG_HAS_SMART_POWER_DOWN) &&
	    !(adapter->flags & FLAG_SMART_POWER_DOWN)) {
3535
		u16 phy_data = 0;
3536 3537
		/*
		 * speed up time to link by disabling smart power down, ignore
3538
		 * the return value of this function because there is nothing
3539 3540
		 * different we would do if it failed
		 */
3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555
		e1e_rphy(hw, IGP02E1000_PHY_POWER_MGMT, &phy_data);
		phy_data &= ~IGP02E1000_PM_SPD;
		e1e_wphy(hw, IGP02E1000_PHY_POWER_MGMT, phy_data);
	}
}

int e1000e_up(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

	/* hardware has been reset, we need to reload some things */
	e1000_configure(adapter);

	clear_bit(__E1000_DOWN, &adapter->state);

3556 3557
	if (adapter->msix_entries)
		e1000_configure_msix(adapter);
3558 3559
	e1000_irq_enable(adapter);

3560
	netif_start_queue(adapter->netdev);
3561

3562
	/* fire a link change interrupt to start the watchdog */
3563 3564 3565 3566 3567
	if (adapter->msix_entries)
		ew32(ICS, E1000_ICS_LSC | E1000_ICR_OTHER);
	else
		ew32(ICS, E1000_ICS_LSC);

3568 3569 3570
	return 0;
}

3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585
static void e1000e_flush_descriptors(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

	if (!(adapter->flags2 & FLAG2_DMA_BURST))
		return;

	/* flush pending descriptor writebacks to memory */
	ew32(TIDV, adapter->tx_int_delay | E1000_TIDV_FPD);
	ew32(RDTR, adapter->rx_int_delay | E1000_RDTR_FPD);

	/* execute the writes immediately */
	e1e_flush();
}

J
Jeff Kirsher 已提交
3586 3587
static void e1000e_update_stats(struct e1000_adapter *adapter);

3588 3589 3590 3591 3592 3593
void e1000e_down(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	struct e1000_hw *hw = &adapter->hw;
	u32 tctl, rctl;

3594 3595 3596 3597
	/*
	 * signal that we're down so the interrupt handler does not
	 * reschedule our watchdog timer
	 */
3598 3599 3600 3601
	set_bit(__E1000_DOWN, &adapter->state);

	/* disable receives in the hardware */
	rctl = er32(RCTL);
3602 3603
	if (!(adapter->flags2 & FLAG2_NO_DISABLE_RX))
		ew32(RCTL, rctl & ~E1000_RCTL_EN);
3604 3605
	/* flush and sleep below */

3606
	netif_stop_queue(netdev);
3607 3608 3609 3610 3611

	/* disable transmits in the hardware */
	tctl = er32(TCTL);
	tctl &= ~E1000_TCTL_EN;
	ew32(TCTL, tctl);
3612

3613 3614
	/* flush both disables and wait for them to finish */
	e1e_flush();
3615
	usleep_range(10000, 20000);
3616 3617 3618 3619 3620 3621 3622

	e1000_irq_disable(adapter);

	del_timer_sync(&adapter->watchdog_timer);
	del_timer_sync(&adapter->phy_info_timer);

	netif_carrier_off(netdev);
J
Jeff Kirsher 已提交
3623 3624 3625 3626 3627

	spin_lock(&adapter->stats64_lock);
	e1000e_update_stats(adapter);
	spin_unlock(&adapter->stats64_lock);

3628
	e1000e_flush_descriptors(adapter);
3629 3630
	e1000_clean_tx_ring(adapter->tx_ring);
	e1000_clean_rx_ring(adapter->rx_ring);
3631

3632 3633 3634
	adapter->link_speed = 0;
	adapter->link_duplex = 0;

3635 3636
	if (!pci_channel_offline(adapter->pdev))
		e1000e_reset(adapter);
3637

3638 3639 3640 3641 3642 3643 3644 3645 3646 3647
	/*
	 * TODO: for power management, we could drop the link and
	 * pci_disable_device here.
	 */
}

void e1000e_reinit_locked(struct e1000_adapter *adapter)
{
	might_sleep();
	while (test_and_set_bit(__E1000_RESETTING, &adapter->state))
3648
		usleep_range(1000, 2000);
3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667
	e1000e_down(adapter);
	e1000e_up(adapter);
	clear_bit(__E1000_RESETTING, &adapter->state);
}

/**
 * e1000_sw_init - Initialize general software structures (struct e1000_adapter)
 * @adapter: board private structure to initialize
 *
 * e1000_sw_init initializes the Adapter private data structure.
 * Fields are initialized based on PCI device information and
 * OS network device settings (MTU size).
 **/
static int __devinit e1000_sw_init(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;

	adapter->rx_buffer_len = ETH_FRAME_LEN + VLAN_HLEN + ETH_FCS_LEN;
	adapter->rx_ps_bsize0 = 128;
3668 3669
	adapter->max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
	adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN;
3670 3671
	adapter->tx_ring_count = E1000_DEFAULT_TXD;
	adapter->rx_ring_count = E1000_DEFAULT_RXD;
3672

J
Jeff Kirsher 已提交
3673 3674
	spin_lock_init(&adapter->stats64_lock);

3675
	e1000e_set_interrupt_capability(adapter);
3676

3677 3678
	if (e1000_alloc_queues(adapter))
		return -ENOMEM;
3679 3680 3681 3682 3683 3684 3685 3686

	/* Explicitly disable IRQ since the NIC can be in any state. */
	e1000_irq_disable(adapter);

	set_bit(__E1000_DOWN, &adapter->state);
	return 0;
}

3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698
/**
 * e1000_intr_msi_test - Interrupt Handler
 * @irq: interrupt number
 * @data: pointer to a network interface device structure
 **/
static irqreturn_t e1000_intr_msi_test(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 icr = er32(ICR);

3699
	e_dbg("icr is %08X\n", icr);
3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725
	if (icr & E1000_ICR_RXSEQ) {
		adapter->flags &= ~FLAG_MSI_TEST_FAILED;
		wmb();
	}

	return IRQ_HANDLED;
}

/**
 * e1000_test_msi_interrupt - Returns 0 for successful test
 * @adapter: board private struct
 *
 * code flow taken from tg3.c
 **/
static int e1000_test_msi_interrupt(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	struct e1000_hw *hw = &adapter->hw;
	int err;

	/* poll_enable hasn't been called yet, so don't need disable */
	/* clear any pending events */
	er32(ICR);

	/* free the real vector and request a test handler */
	e1000_free_irq(adapter);
3726
	e1000e_reset_interrupt_capability(adapter);
3727 3728 3729 3730 3731 3732 3733 3734 3735

	/* Assume that the test fails, if it succeeds then the test
	 * MSI irq handler will unset this flag */
	adapter->flags |= FLAG_MSI_TEST_FAILED;

	err = pci_enable_msi(adapter->pdev);
	if (err)
		goto msi_test_failed;

3736
	err = request_irq(adapter->pdev->irq, e1000_intr_msi_test, 0,
3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756
			  netdev->name, netdev);
	if (err) {
		pci_disable_msi(adapter->pdev);
		goto msi_test_failed;
	}

	wmb();

	e1000_irq_enable(adapter);

	/* fire an unusual interrupt on the test handler */
	ew32(ICS, E1000_ICS_RXSEQ);
	e1e_flush();
	msleep(50);

	e1000_irq_disable(adapter);

	rmb();

	if (adapter->flags & FLAG_MSI_TEST_FAILED) {
3757
		adapter->int_mode = E1000E_INT_MODE_LEGACY;
3758 3759 3760
		e_info("MSI interrupt test failed, using legacy interrupt.\n");
	} else
		e_dbg("MSI interrupt test succeeded!\n");
3761 3762 3763 3764 3765

	free_irq(adapter->pdev->irq, netdev);
	pci_disable_msi(adapter->pdev);

msi_test_failed:
3766
	e1000e_set_interrupt_capability(adapter);
3767
	return e1000_request_irq(adapter);
3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785
}

/**
 * e1000_test_msi - Returns 0 if MSI test succeeds or INTx mode is restored
 * @adapter: board private struct
 *
 * code flow taken from tg3.c, called with e1000 interrupts disabled.
 **/
static int e1000_test_msi(struct e1000_adapter *adapter)
{
	int err;
	u16 pci_cmd;

	if (!(adapter->flags & FLAG_MSI_ENABLED))
		return 0;

	/* disable SERR in case the MSI write causes a master abort */
	pci_read_config_word(adapter->pdev, PCI_COMMAND, &pci_cmd);
3786 3787 3788
	if (pci_cmd & PCI_COMMAND_SERR)
		pci_write_config_word(adapter->pdev, PCI_COMMAND,
				      pci_cmd & ~PCI_COMMAND_SERR);
3789 3790 3791

	err = e1000_test_msi_interrupt(adapter);

3792 3793 3794 3795 3796 3797
	/* re-enable SERR */
	if (pci_cmd & PCI_COMMAND_SERR) {
		pci_read_config_word(adapter->pdev, PCI_COMMAND, &pci_cmd);
		pci_cmd |= PCI_COMMAND_SERR;
		pci_write_config_word(adapter->pdev, PCI_COMMAND, pci_cmd);
	}
3798 3799 3800 3801

	return err;
}

3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817
/**
 * e1000_open - Called when a network interface is made active
 * @netdev: network interface device structure
 *
 * Returns 0 on success, negative value on failure
 *
 * The open entry point is called when a network interface is made
 * active by the system (IFF_UP).  At this point all resources needed
 * for transmit and receive operations are allocated, the interrupt
 * handler is registered with the OS, the watchdog timer is started,
 * and the stack is notified that the interface is ready.
 **/
static int e1000_open(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
3818
	struct pci_dev *pdev = adapter->pdev;
3819 3820 3821 3822 3823 3824
	int err;

	/* disallow open during test */
	if (test_bit(__E1000_TESTING, &adapter->state))
		return -EBUSY;

3825 3826
	pm_runtime_get_sync(&pdev->dev);

3827 3828
	netif_carrier_off(netdev);

3829
	/* allocate transmit descriptors */
3830
	err = e1000e_setup_tx_resources(adapter->tx_ring);
3831 3832 3833 3834
	if (err)
		goto err_setup_tx;

	/* allocate receive descriptors */
3835
	err = e1000e_setup_rx_resources(adapter->rx_ring);
3836 3837 3838
	if (err)
		goto err_setup_rx;

3839 3840 3841 3842 3843
	/*
	 * If AMT is enabled, let the firmware know that the network
	 * interface is now open and reset the part to a known state.
	 */
	if (adapter->flags & FLAG_HAS_AMT) {
3844
		e1000e_get_hw_control(adapter);
3845 3846 3847
		e1000e_reset(adapter);
	}

3848 3849 3850 3851 3852 3853 3854
	e1000e_power_up_phy(adapter);

	adapter->mng_vlan_id = E1000_MNG_VLAN_NONE;
	if ((adapter->hw.mng_cookie.status &
	     E1000_MNG_DHCP_COOKIE_STATUS_VLAN))
		e1000_update_mng_vlan(adapter);

3855 3856
	/* DMA latency requirement to workaround jumbo issue */
	if (adapter->hw.mac.type == e1000_pch2lan)
3857 3858 3859
		pm_qos_add_request(&adapter->netdev->pm_qos_req,
				   PM_QOS_CPU_DMA_LATENCY,
				   PM_QOS_DEFAULT_VALUE);
3860

3861 3862
	/*
	 * before we allocate an interrupt, we must be ready to handle it.
3863 3864
	 * Setting DEBUG_SHIRQ in the kernel makes it fire an interrupt
	 * as soon as we call pci_request_irq, so we have to setup our
3865 3866
	 * clean_rx handler before we do so.
	 */
3867 3868 3869 3870 3871 3872
	e1000_configure(adapter);

	err = e1000_request_irq(adapter);
	if (err)
		goto err_req_irq;

3873 3874 3875 3876 3877
	/*
	 * Work around PCIe errata with MSI interrupts causing some chipsets to
	 * ignore e1000e MSI messages, which means we need to test our MSI
	 * interrupt now
	 */
3878
	if (adapter->int_mode != E1000E_INT_MODE_LEGACY) {
3879 3880 3881 3882 3883 3884 3885
		err = e1000_test_msi(adapter);
		if (err) {
			e_err("Interrupt allocation failed\n");
			goto err_req_irq;
		}
	}

3886 3887 3888 3889 3890 3891 3892
	/* From here on the code is the same as e1000e_up() */
	clear_bit(__E1000_DOWN, &adapter->state);

	napi_enable(&adapter->napi);

	e1000_irq_enable(adapter);

3893
	adapter->tx_hang_recheck = false;
3894
	netif_start_queue(netdev);
3895

3896 3897 3898
	adapter->idle_check = true;
	pm_runtime_put(&pdev->dev);

3899
	/* fire a link status change interrupt to start the watchdog */
3900 3901 3902 3903
	if (adapter->msix_entries)
		ew32(ICS, E1000_ICS_LSC | E1000_ICR_OTHER);
	else
		ew32(ICS, E1000_ICS_LSC);
3904 3905 3906 3907

	return 0;

err_req_irq:
3908
	e1000e_release_hw_control(adapter);
3909
	e1000_power_down_phy(adapter);
3910
	e1000e_free_rx_resources(adapter->rx_ring);
3911
err_setup_rx:
3912
	e1000e_free_tx_resources(adapter->tx_ring);
3913 3914
err_setup_tx:
	e1000e_reset(adapter);
3915
	pm_runtime_put_sync(&pdev->dev);
3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933

	return err;
}

/**
 * e1000_close - Disables a network interface
 * @netdev: network interface device structure
 *
 * Returns 0, this is not allowed to fail
 *
 * The close entry point is called when an interface is de-activated
 * by the OS.  The hardware is still under the drivers control, but
 * needs to be disabled.  A global MAC reset is issued to stop the
 * hardware, and all transmit and receive resources are freed.
 **/
static int e1000_close(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
3934
	struct pci_dev *pdev = adapter->pdev;
3935 3936

	WARN_ON(test_bit(__E1000_RESETTING, &adapter->state));
3937 3938 3939

	pm_runtime_get_sync(&pdev->dev);

3940 3941
	napi_disable(&adapter->napi);

3942 3943 3944 3945
	if (!test_bit(__E1000_DOWN, &adapter->state)) {
		e1000e_down(adapter);
		e1000_free_irq(adapter);
	}
3946 3947
	e1000_power_down_phy(adapter);

3948 3949
	e1000e_free_tx_resources(adapter->tx_ring);
	e1000e_free_rx_resources(adapter->rx_ring);
3950

3951 3952 3953 3954
	/*
	 * kill manageability vlan ID if supported, but not if a vlan with
	 * the same ID is registered on the host OS (let 8021q kill it)
	 */
J
Jeff Kirsher 已提交
3955 3956
	if (adapter->hw.mng_cookie.status &
	    E1000_MNG_DHCP_COOKIE_STATUS_VLAN)
3957 3958
		e1000_vlan_rx_kill_vid(netdev, adapter->mng_vlan_id);

3959 3960 3961 3962
	/*
	 * If AMT is enabled, let the firmware know that the network
	 * interface is now closed
	 */
3963 3964 3965
	if ((adapter->flags & FLAG_HAS_AMT) &&
	    !test_bit(__E1000_TESTING, &adapter->state))
		e1000e_release_hw_control(adapter);
3966

3967
	if (adapter->hw.mac.type == e1000_pch2lan)
3968
		pm_qos_remove_request(&adapter->netdev->pm_qos_req);
3969

3970 3971
	pm_runtime_put_sync(&pdev->dev);

3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997
	return 0;
}
/**
 * e1000_set_mac - Change the Ethernet Address of the NIC
 * @netdev: network interface device structure
 * @p: pointer to an address structure
 *
 * Returns 0 on success, negative on failure
 **/
static int e1000_set_mac(struct net_device *netdev, void *p)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct sockaddr *addr = p;

	if (!is_valid_ether_addr(addr->sa_data))
		return -EADDRNOTAVAIL;

	memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
	memcpy(adapter->hw.mac.addr, addr->sa_data, netdev->addr_len);

	e1000e_rar_set(&adapter->hw, adapter->hw.mac.addr, 0);

	if (adapter->flags & FLAG_RESET_OVERWRITES_LAA) {
		/* activate the work around */
		e1000e_set_laa_state_82571(&adapter->hw, 1);

3998 3999
		/*
		 * Hold a copy of the LAA in RAR[14] This is done so that
4000 4001 4002 4003
		 * between the time RAR[0] gets clobbered  and the time it
		 * gets fixed (in e1000_watchdog), the actual LAA is in one
		 * of the RARs and no incoming packets directed to this port
		 * are dropped. Eventually the LAA will be in RAR[0] and
4004 4005
		 * RAR[14]
		 */
4006 4007 4008 4009 4010 4011 4012 4013
		e1000e_rar_set(&adapter->hw,
			      adapter->hw.mac.addr,
			      adapter->hw.mac.rar_entry_count - 1);
	}

	return 0;
}

4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025
/**
 * e1000e_update_phy_task - work thread to update phy
 * @work: pointer to our work struct
 *
 * this worker thread exists because we must acquire a
 * semaphore to read the phy, which we could msleep while
 * waiting for it, and we can't msleep in a timer.
 **/
static void e1000e_update_phy_task(struct work_struct *work)
{
	struct e1000_adapter *adapter = container_of(work,
					struct e1000_adapter, update_phy_task);
4026 4027 4028 4029

	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

4030 4031 4032
	e1000_get_phy_info(&adapter->hw);
}

4033 4034 4035 4036
/*
 * Need to wait a few seconds after link up to get diagnostic information from
 * the phy
 */
4037 4038 4039
static void e1000_update_phy_info(unsigned long data)
{
	struct e1000_adapter *adapter = (struct e1000_adapter *) data;
4040 4041 4042 4043

	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

4044
	schedule_work(&adapter->update_phy_task);
4045 4046
}

4047 4048 4049
/**
 * e1000e_update_phy_stats - Update the PHY statistics counters
 * @adapter: board private structure
4050 4051
 *
 * Read/clear the upper 16-bit PHY registers and read/accumulate lower
4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066
 **/
static void e1000e_update_phy_stats(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	s32 ret_val;
	u16 phy_data;

	ret_val = hw->phy.ops.acquire(hw);
	if (ret_val)
		return;

	/*
	 * A page set is expensive so check if already on desired page.
	 * If not, set to the page with the PHY status registers.
	 */
4067
	hw->phy.addr = 1;
4068 4069 4070 4071
	ret_val = e1000e_read_phy_reg_mdic(hw, IGP01E1000_PHY_PAGE_SELECT,
					   &phy_data);
	if (ret_val)
		goto release;
4072 4073 4074
	if (phy_data != (HV_STATS_PAGE << IGP_PAGE_SHIFT)) {
		ret_val = hw->phy.ops.set_page(hw,
					       HV_STATS_PAGE << IGP_PAGE_SHIFT);
4075 4076 4077 4078 4079
		if (ret_val)
			goto release;
	}

	/* Single Collision Count */
4080 4081
	hw->phy.ops.read_reg_page(hw, HV_SCC_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_SCC_LOWER, &phy_data);
4082 4083 4084 4085
	if (!ret_val)
		adapter->stats.scc += phy_data;

	/* Excessive Collision Count */
4086 4087
	hw->phy.ops.read_reg_page(hw, HV_ECOL_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_ECOL_LOWER, &phy_data);
4088 4089 4090 4091
	if (!ret_val)
		adapter->stats.ecol += phy_data;

	/* Multiple Collision Count */
4092 4093
	hw->phy.ops.read_reg_page(hw, HV_MCC_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_MCC_LOWER, &phy_data);
4094 4095 4096 4097
	if (!ret_val)
		adapter->stats.mcc += phy_data;

	/* Late Collision Count */
4098 4099
	hw->phy.ops.read_reg_page(hw, HV_LATECOL_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_LATECOL_LOWER, &phy_data);
4100 4101 4102 4103
	if (!ret_val)
		adapter->stats.latecol += phy_data;

	/* Collision Count - also used for adaptive IFS */
4104 4105
	hw->phy.ops.read_reg_page(hw, HV_COLC_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_COLC_LOWER, &phy_data);
4106 4107 4108 4109
	if (!ret_val)
		hw->mac.collision_delta = phy_data;

	/* Defer Count */
4110 4111
	hw->phy.ops.read_reg_page(hw, HV_DC_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_DC_LOWER, &phy_data);
4112 4113 4114 4115
	if (!ret_val)
		adapter->stats.dc += phy_data;

	/* Transmit with no CRS */
4116 4117
	hw->phy.ops.read_reg_page(hw, HV_TNCRS_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_TNCRS_LOWER, &phy_data);
4118 4119 4120 4121 4122 4123 4124
	if (!ret_val)
		adapter->stats.tncrs += phy_data;

release:
	hw->phy.ops.release(hw);
}

4125 4126 4127 4128
/**
 * e1000e_update_stats - Update the board statistics counters
 * @adapter: board private structure
 **/
J
Jeff Kirsher 已提交
4129
static void e1000e_update_stats(struct e1000_adapter *adapter)
4130
{
4131
	struct net_device *netdev = adapter->netdev;
4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145
	struct e1000_hw *hw = &adapter->hw;
	struct pci_dev *pdev = adapter->pdev;

	/*
	 * Prevent stats update while adapter is being reset, or if the pci
	 * connection is down.
	 */
	if (adapter->link_speed == 0)
		return;
	if (pci_channel_offline(pdev))
		return;

	adapter->stats.crcerrs += er32(CRCERRS);
	adapter->stats.gprc += er32(GPRC);
4146 4147
	adapter->stats.gorc += er32(GORCL);
	er32(GORCH); /* Clear gorc */
4148 4149 4150 4151 4152
	adapter->stats.bprc += er32(BPRC);
	adapter->stats.mprc += er32(MPRC);
	adapter->stats.roc += er32(ROC);

	adapter->stats.mpc += er32(MPC);
4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171

	/* Half-duplex statistics */
	if (adapter->link_duplex == HALF_DUPLEX) {
		if (adapter->flags2 & FLAG2_HAS_PHY_STATS) {
			e1000e_update_phy_stats(adapter);
		} else {
			adapter->stats.scc += er32(SCC);
			adapter->stats.ecol += er32(ECOL);
			adapter->stats.mcc += er32(MCC);
			adapter->stats.latecol += er32(LATECOL);
			adapter->stats.dc += er32(DC);

			hw->mac.collision_delta = er32(COLC);

			if ((hw->mac.type != e1000_82574) &&
			    (hw->mac.type != e1000_82583))
				adapter->stats.tncrs += er32(TNCRS);
		}
		adapter->stats.colc += hw->mac.collision_delta;
4172
	}
4173

4174 4175 4176 4177 4178
	adapter->stats.xonrxc += er32(XONRXC);
	adapter->stats.xontxc += er32(XONTXC);
	adapter->stats.xoffrxc += er32(XOFFRXC);
	adapter->stats.xofftxc += er32(XOFFTXC);
	adapter->stats.gptc += er32(GPTC);
4179 4180
	adapter->stats.gotc += er32(GOTCL);
	er32(GOTCH); /* Clear gotc */
4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198
	adapter->stats.rnbc += er32(RNBC);
	adapter->stats.ruc += er32(RUC);

	adapter->stats.mptc += er32(MPTC);
	adapter->stats.bptc += er32(BPTC);

	/* used for adaptive IFS */

	hw->mac.tx_packet_delta = er32(TPT);
	adapter->stats.tpt += hw->mac.tx_packet_delta;

	adapter->stats.algnerrc += er32(ALGNERRC);
	adapter->stats.rxerrc += er32(RXERRC);
	adapter->stats.cexterr += er32(CEXTERR);
	adapter->stats.tsctc += er32(TSCTC);
	adapter->stats.tsctfc += er32(TSCTFC);

	/* Fill out the OS statistics structure */
4199 4200
	netdev->stats.multicast = adapter->stats.mprc;
	netdev->stats.collisions = adapter->stats.colc;
4201 4202 4203

	/* Rx Errors */

4204 4205 4206 4207
	/*
	 * RLEC on some newer hardware can be incorrect so build
	 * our own version based on RUC and ROC
	 */
4208
	netdev->stats.rx_errors = adapter->stats.rxerrc +
4209 4210 4211
		adapter->stats.crcerrs + adapter->stats.algnerrc +
		adapter->stats.ruc + adapter->stats.roc +
		adapter->stats.cexterr;
4212
	netdev->stats.rx_length_errors = adapter->stats.ruc +
4213
					      adapter->stats.roc;
4214 4215 4216
	netdev->stats.rx_crc_errors = adapter->stats.crcerrs;
	netdev->stats.rx_frame_errors = adapter->stats.algnerrc;
	netdev->stats.rx_missed_errors = adapter->stats.mpc;
4217 4218

	/* Tx Errors */
4219
	netdev->stats.tx_errors = adapter->stats.ecol +
4220
				       adapter->stats.latecol;
4221 4222 4223
	netdev->stats.tx_aborted_errors = adapter->stats.ecol;
	netdev->stats.tx_window_errors = adapter->stats.latecol;
	netdev->stats.tx_carrier_errors = adapter->stats.tncrs;
4224 4225 4226 4227 4228 4229 4230 4231 4232

	/* Tx Dropped needs to be maintained elsewhere */

	/* Management Stats */
	adapter->stats.mgptc += er32(MGTPTC);
	adapter->stats.mgprc += er32(MGTPRC);
	adapter->stats.mgpdc += er32(MGTPDC);
}

4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243
/**
 * e1000_phy_read_status - Update the PHY register status snapshot
 * @adapter: board private structure
 **/
static void e1000_phy_read_status(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_phy_regs *phy = &adapter->phy_regs;

	if ((er32(STATUS) & E1000_STATUS_LU) &&
	    (adapter->hw.phy.media_type == e1000_media_type_copper)) {
4244 4245
		int ret_val;

4246 4247 4248 4249 4250 4251 4252 4253 4254
		ret_val  = e1e_rphy(hw, PHY_CONTROL, &phy->bmcr);
		ret_val |= e1e_rphy(hw, PHY_STATUS, &phy->bmsr);
		ret_val |= e1e_rphy(hw, PHY_AUTONEG_ADV, &phy->advertise);
		ret_val |= e1e_rphy(hw, PHY_LP_ABILITY, &phy->lpa);
		ret_val |= e1e_rphy(hw, PHY_AUTONEG_EXP, &phy->expansion);
		ret_val |= e1e_rphy(hw, PHY_1000T_CTRL, &phy->ctrl1000);
		ret_val |= e1e_rphy(hw, PHY_1000T_STATUS, &phy->stat1000);
		ret_val |= e1e_rphy(hw, PHY_EXT_STATUS, &phy->estatus);
		if (ret_val)
4255
			e_warn("Error reading PHY register\n");
4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274
	} else {
		/*
		 * Do not read PHY registers if link is not up
		 * Set values to typical power-on defaults
		 */
		phy->bmcr = (BMCR_SPEED1000 | BMCR_ANENABLE | BMCR_FULLDPLX);
		phy->bmsr = (BMSR_100FULL | BMSR_100HALF | BMSR_10FULL |
			     BMSR_10HALF | BMSR_ESTATEN | BMSR_ANEGCAPABLE |
			     BMSR_ERCAP);
		phy->advertise = (ADVERTISE_PAUSE_ASYM | ADVERTISE_PAUSE_CAP |
				  ADVERTISE_ALL | ADVERTISE_CSMA);
		phy->lpa = 0;
		phy->expansion = EXPANSION_ENABLENPAGE;
		phy->ctrl1000 = ADVERTISE_1000FULL;
		phy->stat1000 = 0;
		phy->estatus = (ESTATUS_1000_TFULL | ESTATUS_1000_THALF);
	}
}

4275 4276 4277 4278 4279
static void e1000_print_link_info(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl = er32(CTRL);

4280
	/* Link status message must follow this format for user tools */
4281 4282 4283 4284 4285 4286 4287
	printk(KERN_INFO "e1000e: %s NIC Link is Up %d Mbps %s Duplex, Flow Control: %s\n",
		adapter->netdev->name,
		adapter->link_speed,
		adapter->link_duplex == FULL_DUPLEX ? "Full" : "Half",
		(ctrl & E1000_CTRL_TFCE) && (ctrl & E1000_CTRL_RFCE) ? "Rx/Tx" :
		(ctrl & E1000_CTRL_RFCE) ? "Rx" :
		(ctrl & E1000_CTRL_TFCE) ? "Tx" : "None");
4288 4289
}

4290
static bool e1000e_has_link(struct e1000_adapter *adapter)
4291 4292
{
	struct e1000_hw *hw = &adapter->hw;
4293
	bool link_active = false;
4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307
	s32 ret_val = 0;

	/*
	 * get_link_status is set on LSC (link status) interrupt or
	 * Rx sequence error interrupt.  get_link_status will stay
	 * false until the check_for_link establishes link
	 * for copper adapters ONLY
	 */
	switch (hw->phy.media_type) {
	case e1000_media_type_copper:
		if (hw->mac.get_link_status) {
			ret_val = hw->mac.ops.check_for_link(hw);
			link_active = !hw->mac.get_link_status;
		} else {
4308
			link_active = true;
4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326
		}
		break;
	case e1000_media_type_fiber:
		ret_val = hw->mac.ops.check_for_link(hw);
		link_active = !!(er32(STATUS) & E1000_STATUS_LU);
		break;
	case e1000_media_type_internal_serdes:
		ret_val = hw->mac.ops.check_for_link(hw);
		link_active = adapter->hw.mac.serdes_has_link;
		break;
	default:
	case e1000_media_type_unknown:
		break;
	}

	if ((ret_val == E1000_ERR_PHY) && (hw->phy.type == e1000_phy_igp_3) &&
	    (er32(CTRL) & E1000_PHY_CTRL_GBE_DISABLE)) {
		/* See e1000_kmrn_lock_loss_workaround_ich8lan() */
4327
		e_info("Gigabit has been disabled, downgrading speed\n");
4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344
	}

	return link_active;
}

static void e1000e_enable_receives(struct e1000_adapter *adapter)
{
	/* make sure the receive unit is started */
	if ((adapter->flags & FLAG_RX_NEEDS_RESTART) &&
	    (adapter->flags & FLAG_RX_RESTART_NOW)) {
		struct e1000_hw *hw = &adapter->hw;
		u32 rctl = er32(RCTL);
		ew32(RCTL, rctl | E1000_RCTL_EN);
		adapter->flags &= ~FLAG_RX_RESTART_NOW;
	}
}

4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363
static void e1000e_check_82574_phy_workaround(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

	/*
	 * With 82574 controllers, PHY needs to be checked periodically
	 * for hung state and reset, if two calls return true
	 */
	if (e1000_check_phy_82574(hw))
		adapter->phy_hang_count++;
	else
		adapter->phy_hang_count = 0;

	if (adapter->phy_hang_count > 1) {
		adapter->phy_hang_count = 0;
		schedule_work(&adapter->reset_task);
	}
}

4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383
/**
 * e1000_watchdog - Timer Call-back
 * @data: pointer to adapter cast into an unsigned long
 **/
static void e1000_watchdog(unsigned long data)
{
	struct e1000_adapter *adapter = (struct e1000_adapter *) data;

	/* Do the rest outside of interrupt context */
	schedule_work(&adapter->watchdog_task);

	/* TODO: make this use queue_delayed_work() */
}

static void e1000_watchdog_task(struct work_struct *work)
{
	struct e1000_adapter *adapter = container_of(work,
					struct e1000_adapter, watchdog_task);
	struct net_device *netdev = adapter->netdev;
	struct e1000_mac_info *mac = &adapter->hw.mac;
B
Bruce Allan 已提交
4384
	struct e1000_phy_info *phy = &adapter->hw.phy;
4385 4386 4387 4388
	struct e1000_ring *tx_ring = adapter->tx_ring;
	struct e1000_hw *hw = &adapter->hw;
	u32 link, tctl;

4389 4390 4391
	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

4392
	link = e1000e_has_link(adapter);
4393
	if ((netif_carrier_ok(netdev)) && link) {
4394 4395 4396
		/* Cancel scheduled suspend requests. */
		pm_runtime_resume(netdev->dev.parent);

4397
		e1000e_enable_receives(adapter);
4398 4399 4400 4401 4402 4403 4404 4405 4406
		goto link_up;
	}

	if ((e1000e_enable_tx_pkt_filtering(hw)) &&
	    (adapter->mng_vlan_id != adapter->hw.mng_cookie.vlan_id))
		e1000_update_mng_vlan(adapter);

	if (link) {
		if (!netif_carrier_ok(netdev)) {
4407
			bool txb2b = true;
4408 4409 4410 4411

			/* Cancel scheduled suspend requests. */
			pm_runtime_resume(netdev->dev.parent);

4412
			/* update snapshot of PHY registers on LSC */
4413
			e1000_phy_read_status(adapter);
4414 4415 4416 4417
			mac->ops.get_link_up_info(&adapter->hw,
						   &adapter->link_speed,
						   &adapter->link_duplex);
			e1000_print_link_info(adapter);
4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432
			/*
			 * On supported PHYs, check for duplex mismatch only
			 * if link has autonegotiated at 10/100 half
			 */
			if ((hw->phy.type == e1000_phy_igp_3 ||
			     hw->phy.type == e1000_phy_bm) &&
			    (hw->mac.autoneg == true) &&
			    (adapter->link_speed == SPEED_10 ||
			     adapter->link_speed == SPEED_100) &&
			    (adapter->link_duplex == HALF_DUPLEX)) {
				u16 autoneg_exp;

				e1e_rphy(hw, PHY_AUTONEG_EXP, &autoneg_exp);

				if (!(autoneg_exp & NWAY_ER_LP_NWAY_CAPS))
4433
					e_info("Autonegotiated half duplex but link partner cannot autoneg.  Try forcing full duplex if link gets many collisions.\n");
4434 4435
			}

4436
			/* adjust timeout factor according to speed/duplex */
4437 4438 4439
			adapter->tx_timeout_factor = 1;
			switch (adapter->link_speed) {
			case SPEED_10:
4440
				txb2b = false;
4441
				adapter->tx_timeout_factor = 16;
4442 4443
				break;
			case SPEED_100:
4444
				txb2b = false;
4445
				adapter->tx_timeout_factor = 10;
4446 4447 4448
				break;
			}

4449 4450 4451 4452
			/*
			 * workaround: re-program speed mode bit after
			 * link-up event
			 */
4453 4454 4455
			if ((adapter->flags & FLAG_TARC_SPEED_MODE_BIT) &&
			    !txb2b) {
				u32 tarc0;
4456
				tarc0 = er32(TARC(0));
4457
				tarc0 &= ~SPEED_MODE_BIT;
4458
				ew32(TARC(0), tarc0);
4459 4460
			}

4461 4462 4463 4464
			/*
			 * disable TSO for pcie and 10/100 speeds, to avoid
			 * some hardware issues
			 */
4465 4466 4467 4468
			if (!(adapter->flags & FLAG_TSO_FORCE)) {
				switch (adapter->link_speed) {
				case SPEED_10:
				case SPEED_100:
4469
					e_info("10/100 speed: disabling TSO\n");
4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482
					netdev->features &= ~NETIF_F_TSO;
					netdev->features &= ~NETIF_F_TSO6;
					break;
				case SPEED_1000:
					netdev->features |= NETIF_F_TSO;
					netdev->features |= NETIF_F_TSO6;
					break;
				default:
					/* oops */
					break;
				}
			}

4483 4484 4485 4486
			/*
			 * enable transmits in the hardware, need to do this
			 * after setting TARC(0)
			 */
4487 4488 4489 4490
			tctl = er32(TCTL);
			tctl |= E1000_TCTL_EN;
			ew32(TCTL, tctl);

B
Bruce Allan 已提交
4491 4492 4493 4494 4495 4496 4497
                        /*
			 * Perform any post-link-up configuration before
			 * reporting link up.
			 */
			if (phy->ops.cfg_on_link_up)
				phy->ops.cfg_on_link_up(hw);

4498 4499 4500 4501 4502 4503 4504 4505 4506 4507
			netif_carrier_on(netdev);

			if (!test_bit(__E1000_DOWN, &adapter->state))
				mod_timer(&adapter->phy_info_timer,
					  round_jiffies(jiffies + 2 * HZ));
		}
	} else {
		if (netif_carrier_ok(netdev)) {
			adapter->link_speed = 0;
			adapter->link_duplex = 0;
4508 4509 4510
			/* Link status message must follow this format */
			printk(KERN_INFO "e1000e: %s NIC Link is Down\n",
			       adapter->netdev->name);
4511 4512 4513 4514 4515 4516 4517
			netif_carrier_off(netdev);
			if (!test_bit(__E1000_DOWN, &adapter->state))
				mod_timer(&adapter->phy_info_timer,
					  round_jiffies(jiffies + 2 * HZ));

			if (adapter->flags & FLAG_RX_NEEDS_RESTART)
				schedule_work(&adapter->reset_task);
4518 4519 4520
			else
				pm_schedule_suspend(netdev->dev.parent,
							LINK_TIMEOUT);
4521 4522 4523 4524
		}
	}

link_up:
J
Jeff Kirsher 已提交
4525
	spin_lock(&adapter->stats64_lock);
4526 4527 4528 4529 4530 4531 4532
	e1000e_update_stats(adapter);

	mac->tx_packet_delta = adapter->stats.tpt - adapter->tpt_old;
	adapter->tpt_old = adapter->stats.tpt;
	mac->collision_delta = adapter->stats.colc - adapter->colc_old;
	adapter->colc_old = adapter->stats.colc;

4533 4534 4535 4536
	adapter->gorc = adapter->stats.gorc - adapter->gorc_old;
	adapter->gorc_old = adapter->stats.gorc;
	adapter->gotc = adapter->stats.gotc - adapter->gotc_old;
	adapter->gotc_old = adapter->stats.gotc;
4537
	spin_unlock(&adapter->stats64_lock);
4538 4539 4540

	e1000e_update_adaptive(&adapter->hw);

4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551
	if (!netif_carrier_ok(netdev) &&
	    (e1000_desc_unused(tx_ring) + 1 < tx_ring->count)) {
		/*
		 * We've lost link, so the controller stops DMA,
		 * but we've got queued Tx work that's never going
		 * to get done, so reset controller to flush Tx.
		 * (Do the reset outside of interrupt context).
		 */
		schedule_work(&adapter->reset_task);
		/* return immediately since reset is imminent */
		return;
4552 4553
	}

4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569
	/* Simple mode for Interrupt Throttle Rate (ITR) */
	if (adapter->itr_setting == 4) {
		/*
		 * Symmetric Tx/Rx gets a reduced ITR=2000;
		 * Total asymmetrical Tx or Rx gets ITR=8000;
		 * everyone else is between 2000-8000.
		 */
		u32 goc = (adapter->gotc + adapter->gorc) / 10000;
		u32 dif = (adapter->gotc > adapter->gorc ?
			    adapter->gotc - adapter->gorc :
			    adapter->gorc - adapter->gotc) / 10000;
		u32 itr = goc > 0 ? (dif * 6000 / goc + 2000) : 8000;

		ew32(ITR, 1000000000 / (itr * 256));
	}

4570
	/* Cause software interrupt to ensure Rx ring is cleaned */
4571 4572 4573 4574
	if (adapter->msix_entries)
		ew32(ICS, adapter->rx_ring->ims_val);
	else
		ew32(ICS, E1000_ICS_RXDMT0);
4575

4576 4577 4578
	/* flush pending descriptors to memory before detecting Tx hang */
	e1000e_flush_descriptors(adapter);

4579
	/* Force detection of hung controller every watchdog period */
4580
	adapter->detect_tx_hung = true;
4581

4582 4583 4584 4585
	/*
	 * With 82571 controllers, LAA may be overwritten due to controller
	 * reset from the other port. Set the appropriate LAA in RAR[0]
	 */
4586 4587 4588
	if (e1000e_get_laa_state_82571(hw))
		e1000e_rar_set(hw, adapter->hw.mac.addr, 0);

4589 4590 4591
	if (adapter->flags2 & FLAG2_CHECK_PHY_HANG)
		e1000e_check_82574_phy_workaround(adapter);

4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604
	/* Reset the timer */
	if (!test_bit(__E1000_DOWN, &adapter->state))
		mod_timer(&adapter->watchdog_timer,
			  round_jiffies(jiffies + 2 * HZ));
}

#define E1000_TX_FLAGS_CSUM		0x00000001
#define E1000_TX_FLAGS_VLAN		0x00000002
#define E1000_TX_FLAGS_TSO		0x00000004
#define E1000_TX_FLAGS_IPV4		0x00000008
#define E1000_TX_FLAGS_VLAN_MASK	0xffff0000
#define E1000_TX_FLAGS_VLAN_SHIFT	16

4605
static int e1000_tso(struct e1000_ring *tx_ring, struct sk_buff *skb)
4606 4607 4608 4609 4610 4611 4612 4613
{
	struct e1000_context_desc *context_desc;
	struct e1000_buffer *buffer_info;
	unsigned int i;
	u32 cmd_length = 0;
	u16 ipcse = 0, tucse, mss;
	u8 ipcss, ipcso, tucss, tucso, hdr_len;

4614 4615
	if (!skb_is_gso(skb))
		return 0;
4616

4617
	if (skb_header_cloned(skb)) {
4618 4619
		int err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);

4620 4621
		if (err)
			return err;
4622 4623
	}

4624 4625 4626 4627 4628 4629 4630 4631 4632 4633
	hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
	mss = skb_shinfo(skb)->gso_size;
	if (skb->protocol == htons(ETH_P_IP)) {
		struct iphdr *iph = ip_hdr(skb);
		iph->tot_len = 0;
		iph->check = 0;
		tcp_hdr(skb)->check = ~csum_tcpudp_magic(iph->saddr, iph->daddr,
		                                         0, IPPROTO_TCP, 0);
		cmd_length = E1000_TXD_CMD_IP;
		ipcse = skb_transport_offset(skb) - 1;
4634
	} else if (skb_is_gso_v6(skb)) {
4635 4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672
		ipv6_hdr(skb)->payload_len = 0;
		tcp_hdr(skb)->check = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
		                                       &ipv6_hdr(skb)->daddr,
		                                       0, IPPROTO_TCP, 0);
		ipcse = 0;
	}
	ipcss = skb_network_offset(skb);
	ipcso = (void *)&(ip_hdr(skb)->check) - (void *)skb->data;
	tucss = skb_transport_offset(skb);
	tucso = (void *)&(tcp_hdr(skb)->check) - (void *)skb->data;
	tucse = 0;

	cmd_length |= (E1000_TXD_CMD_DEXT | E1000_TXD_CMD_TSE |
	               E1000_TXD_CMD_TCP | (skb->len - (hdr_len)));

	i = tx_ring->next_to_use;
	context_desc = E1000_CONTEXT_DESC(*tx_ring, i);
	buffer_info = &tx_ring->buffer_info[i];

	context_desc->lower_setup.ip_fields.ipcss  = ipcss;
	context_desc->lower_setup.ip_fields.ipcso  = ipcso;
	context_desc->lower_setup.ip_fields.ipcse  = cpu_to_le16(ipcse);
	context_desc->upper_setup.tcp_fields.tucss = tucss;
	context_desc->upper_setup.tcp_fields.tucso = tucso;
	context_desc->upper_setup.tcp_fields.tucse = cpu_to_le16(tucse);
	context_desc->tcp_seg_setup.fields.mss     = cpu_to_le16(mss);
	context_desc->tcp_seg_setup.fields.hdr_len = hdr_len;
	context_desc->cmd_and_length = cpu_to_le32(cmd_length);

	buffer_info->time_stamp = jiffies;
	buffer_info->next_to_watch = i;

	i++;
	if (i == tx_ring->count)
		i = 0;
	tx_ring->next_to_use = i;

	return 1;
4673 4674
}

4675
static bool e1000_tx_csum(struct e1000_ring *tx_ring, struct sk_buff *skb)
4676
{
4677
	struct e1000_adapter *adapter = tx_ring->adapter;
4678 4679 4680 4681
	struct e1000_context_desc *context_desc;
	struct e1000_buffer *buffer_info;
	unsigned int i;
	u8 css;
4682
	u32 cmd_len = E1000_TXD_CMD_DEXT;
4683
	__be16 protocol;
4684

4685 4686
	if (skb->ip_summed != CHECKSUM_PARTIAL)
		return 0;
4687

4688 4689 4690 4691 4692
	if (skb->protocol == cpu_to_be16(ETH_P_8021Q))
		protocol = vlan_eth_hdr(skb)->h_vlan_encapsulated_proto;
	else
		protocol = skb->protocol;

A
Arthur Jones 已提交
4693
	switch (protocol) {
4694
	case cpu_to_be16(ETH_P_IP):
4695 4696 4697
		if (ip_hdr(skb)->protocol == IPPROTO_TCP)
			cmd_len |= E1000_TXD_CMD_TCP;
		break;
4698
	case cpu_to_be16(ETH_P_IPV6):
4699 4700 4701 4702 4703 4704
		/* XXX not handling all IPV6 headers */
		if (ipv6_hdr(skb)->nexthdr == IPPROTO_TCP)
			cmd_len |= E1000_TXD_CMD_TCP;
		break;
	default:
		if (unlikely(net_ratelimit()))
4705 4706
			e_warn("checksum_partial proto=%x!\n",
			       be16_to_cpu(protocol));
4707
		break;
4708 4709
	}

4710
	css = skb_checksum_start_offset(skb);
4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732

	i = tx_ring->next_to_use;
	buffer_info = &tx_ring->buffer_info[i];
	context_desc = E1000_CONTEXT_DESC(*tx_ring, i);

	context_desc->lower_setup.ip_config = 0;
	context_desc->upper_setup.tcp_fields.tucss = css;
	context_desc->upper_setup.tcp_fields.tucso =
				css + skb->csum_offset;
	context_desc->upper_setup.tcp_fields.tucse = 0;
	context_desc->tcp_seg_setup.data = 0;
	context_desc->cmd_and_length = cpu_to_le32(cmd_len);

	buffer_info->time_stamp = jiffies;
	buffer_info->next_to_watch = i;

	i++;
	if (i == tx_ring->count)
		i = 0;
	tx_ring->next_to_use = i;

	return 1;
4733 4734 4735 4736 4737
}

#define E1000_MAX_PER_TXD	8192
#define E1000_MAX_TXD_PWR	12

4738 4739 4740
static int e1000_tx_map(struct e1000_ring *tx_ring, struct sk_buff *skb,
			unsigned int first, unsigned int max_per_txd,
			unsigned int nr_frags, unsigned int mss)
4741
{
4742
	struct e1000_adapter *adapter = tx_ring->adapter;
4743
	struct pci_dev *pdev = adapter->pdev;
4744
	struct e1000_buffer *buffer_info;
J
Jesse Brandeburg 已提交
4745
	unsigned int len = skb_headlen(skb);
4746
	unsigned int offset = 0, size, count = 0, i;
4747
	unsigned int f, bytecount, segs;
4748 4749 4750 4751

	i = tx_ring->next_to_use;

	while (len) {
4752
		buffer_info = &tx_ring->buffer_info[i];
4753 4754 4755 4756 4757
		size = min(len, max_per_txd);

		buffer_info->length = size;
		buffer_info->time_stamp = jiffies;
		buffer_info->next_to_watch = i;
4758 4759
		buffer_info->dma = dma_map_single(&pdev->dev,
						  skb->data + offset,
4760
						  size, DMA_TO_DEVICE);
4761
		buffer_info->mapped_as_page = false;
4762
		if (dma_mapping_error(&pdev->dev, buffer_info->dma))
4763
			goto dma_error;
4764 4765 4766

		len -= size;
		offset += size;
4767
		count++;
4768 4769 4770 4771 4772 4773

		if (len) {
			i++;
			if (i == tx_ring->count)
				i = 0;
		}
4774 4775 4776
	}

	for (f = 0; f < nr_frags; f++) {
E
Eric Dumazet 已提交
4777
		const struct skb_frag_struct *frag;
4778 4779

		frag = &skb_shinfo(skb)->frags[f];
E
Eric Dumazet 已提交
4780
		len = skb_frag_size(frag);
4781
		offset = 0;
4782 4783

		while (len) {
4784 4785 4786 4787
			i++;
			if (i == tx_ring->count)
				i = 0;

4788 4789 4790 4791 4792 4793
			buffer_info = &tx_ring->buffer_info[i];
			size = min(len, max_per_txd);

			buffer_info->length = size;
			buffer_info->time_stamp = jiffies;
			buffer_info->next_to_watch = i;
4794 4795
			buffer_info->dma = skb_frag_dma_map(&pdev->dev, frag,
						offset, size, DMA_TO_DEVICE);
4796
			buffer_info->mapped_as_page = true;
4797
			if (dma_mapping_error(&pdev->dev, buffer_info->dma))
4798
				goto dma_error;
4799 4800 4801 4802 4803 4804 4805

			len -= size;
			offset += size;
			count++;
		}
	}

4806
	segs = skb_shinfo(skb)->gso_segs ? : 1;
4807 4808 4809
	/* multiply data chunks by size of headers */
	bytecount = ((segs - 1) * skb_headlen(skb)) + skb->len;

4810
	tx_ring->buffer_info[i].skb = skb;
4811 4812
	tx_ring->buffer_info[i].segs = segs;
	tx_ring->buffer_info[i].bytecount = bytecount;
4813 4814 4815
	tx_ring->buffer_info[first].next_to_watch = i;

	return count;
4816 4817

dma_error:
4818
	dev_err(&pdev->dev, "Tx DMA map failed\n");
4819
	buffer_info->dma = 0;
4820
	if (count)
4821
		count--;
4822 4823

	while (count--) {
4824
		if (i == 0)
4825
			i += tx_ring->count;
4826
		i--;
4827
		buffer_info = &tx_ring->buffer_info[i];
4828
		e1000_put_txbuf(tx_ring, buffer_info);
4829 4830 4831
	}

	return 0;
4832 4833
}

4834
static void e1000_tx_queue(struct e1000_ring *tx_ring, int tx_flags, int count)
4835
{
4836
	struct e1000_adapter *adapter = tx_ring->adapter;
4837 4838 4839 4840 4841 4842 4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862
	struct e1000_tx_desc *tx_desc = NULL;
	struct e1000_buffer *buffer_info;
	u32 txd_upper = 0, txd_lower = E1000_TXD_CMD_IFCS;
	unsigned int i;

	if (tx_flags & E1000_TX_FLAGS_TSO) {
		txd_lower |= E1000_TXD_CMD_DEXT | E1000_TXD_DTYP_D |
			     E1000_TXD_CMD_TSE;
		txd_upper |= E1000_TXD_POPTS_TXSM << 8;

		if (tx_flags & E1000_TX_FLAGS_IPV4)
			txd_upper |= E1000_TXD_POPTS_IXSM << 8;
	}

	if (tx_flags & E1000_TX_FLAGS_CSUM) {
		txd_lower |= E1000_TXD_CMD_DEXT | E1000_TXD_DTYP_D;
		txd_upper |= E1000_TXD_POPTS_TXSM << 8;
	}

	if (tx_flags & E1000_TX_FLAGS_VLAN) {
		txd_lower |= E1000_TXD_CMD_VLE;
		txd_upper |= (tx_flags & E1000_TX_FLAGS_VLAN_MASK);
	}

	i = tx_ring->next_to_use;

4863
	do {
4864 4865 4866 4867 4868 4869 4870 4871 4872 4873
		buffer_info = &tx_ring->buffer_info[i];
		tx_desc = E1000_TX_DESC(*tx_ring, i);
		tx_desc->buffer_addr = cpu_to_le64(buffer_info->dma);
		tx_desc->lower.data =
			cpu_to_le32(txd_lower | buffer_info->length);
		tx_desc->upper.data = cpu_to_le32(txd_upper);

		i++;
		if (i == tx_ring->count)
			i = 0;
4874
	} while (--count > 0);
4875 4876 4877

	tx_desc->lower.data |= cpu_to_le32(adapter->txd_cmd);

4878 4879
	/*
	 * Force memory writes to complete before letting h/w
4880 4881
	 * know there are new descriptors to fetch.  (Only
	 * applicable for weak-ordered memory model archs,
4882 4883
	 * such as IA-64).
	 */
4884 4885 4886
	wmb();

	tx_ring->next_to_use = i;
4887 4888

	if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
4889
		e1000e_update_tdt_wa(tx_ring, i);
4890
	else
4891
		writel(i, tx_ring->tail);
4892

4893 4894 4895 4896
	/*
	 * we need this if more than one processor can write to our tail
	 * at a time, it synchronizes IO on IA64/Altix systems
	 */
4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907
	mmiowb();
}

#define MINIMUM_DHCP_PACKET_SIZE 282
static int e1000_transfer_dhcp_info(struct e1000_adapter *adapter,
				    struct sk_buff *skb)
{
	struct e1000_hw *hw =  &adapter->hw;
	u16 length, offset;

	if (vlan_tx_tag_present(skb)) {
4908 4909
		if (!((vlan_tx_tag_get(skb) == adapter->hw.mng_cookie.vlan_id) &&
		    (adapter->hw.mng_cookie.status &
4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938
			E1000_MNG_DHCP_COOKIE_STATUS_VLAN)))
			return 0;
	}

	if (skb->len <= MINIMUM_DHCP_PACKET_SIZE)
		return 0;

	if (((struct ethhdr *) skb->data)->h_proto != htons(ETH_P_IP))
		return 0;

	{
		const struct iphdr *ip = (struct iphdr *)((u8 *)skb->data+14);
		struct udphdr *udp;

		if (ip->protocol != IPPROTO_UDP)
			return 0;

		udp = (struct udphdr *)((u8 *)ip + (ip->ihl << 2));
		if (ntohs(udp->dest) != 67)
			return 0;

		offset = (u8 *)udp + 8 - skb->data;
		length = skb->len - offset;
		return e1000e_mng_write_dhcp_info(hw, (u8 *)udp + 8, length);
	}

	return 0;
}

4939
static int __e1000_maybe_stop_tx(struct e1000_ring *tx_ring, int size)
4940
{
4941
	struct e1000_adapter *adapter = tx_ring->adapter;
4942

4943
	netif_stop_queue(adapter->netdev);
4944 4945
	/*
	 * Herbert's original patch had:
4946
	 *  smp_mb__after_netif_stop_queue();
4947 4948
	 * but since that doesn't exist yet, just open code it.
	 */
4949 4950
	smp_mb();

4951 4952 4953 4954
	/*
	 * We need to check again in a case another CPU has just
	 * made room available.
	 */
4955
	if (e1000_desc_unused(tx_ring) < size)
4956 4957 4958
		return -EBUSY;

	/* A reprieve! */
4959
	netif_start_queue(adapter->netdev);
4960 4961 4962 4963
	++adapter->restart_queue;
	return 0;
}

4964
static int e1000_maybe_stop_tx(struct e1000_ring *tx_ring, int size)
4965
{
4966
	if (e1000_desc_unused(tx_ring) >= size)
4967
		return 0;
4968
	return __e1000_maybe_stop_tx(tx_ring, size);
4969 4970 4971
}

#define TXD_USE_COUNT(S, X) (((S) >> (X)) + 1 )
4972 4973
static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
				    struct net_device *netdev)
4974 4975 4976 4977 4978 4979 4980
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_ring *tx_ring = adapter->tx_ring;
	unsigned int first;
	unsigned int max_per_txd = E1000_MAX_PER_TXD;
	unsigned int max_txd_pwr = E1000_MAX_TXD_PWR;
	unsigned int tx_flags = 0;
E
Eric Dumazet 已提交
4981
	unsigned int len = skb_headlen(skb);
4982 4983
	unsigned int nr_frags;
	unsigned int mss;
4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996 4997 4998
	int count = 0;
	int tso;
	unsigned int f;

	if (test_bit(__E1000_DOWN, &adapter->state)) {
		dev_kfree_skb_any(skb);
		return NETDEV_TX_OK;
	}

	if (skb->len <= 0) {
		dev_kfree_skb_any(skb);
		return NETDEV_TX_OK;
	}

	mss = skb_shinfo(skb)->gso_size;
4999 5000
	/*
	 * The controller does a simple calculation to
5001 5002 5003 5004
	 * make sure there is enough room in the FIFO before
	 * initiating the DMA for each buffer.  The calc is:
	 * 4 = ceil(buffer len/mss).  To make sure we don't
	 * overrun the FIFO, adjust the max buffer len if mss
5005 5006
	 * drops.
	 */
5007 5008 5009 5010 5011
	if (mss) {
		u8 hdr_len;
		max_per_txd = min(mss << 2, max_per_txd);
		max_txd_pwr = fls(max_per_txd) - 1;

5012 5013 5014 5015 5016
		/*
		 * TSO Workaround for 82571/2/3 Controllers -- if skb->data
		 * points to just header, pull a few bytes of payload from
		 * frags into skb->data
		 */
5017
		hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
5018 5019 5020 5021
		/*
		 * we do this workaround for ES2LAN, but it is un-necessary,
		 * avoiding it could save a lot of cycles
		 */
5022
		if (skb->data_len && (hdr_len == len)) {
5023 5024 5025 5026
			unsigned int pull_size;

			pull_size = min((unsigned int)4, skb->data_len);
			if (!__pskb_pull_tail(skb, pull_size)) {
5027
				e_err("__pskb_pull_tail failed.\n");
5028 5029 5030
				dev_kfree_skb_any(skb);
				return NETDEV_TX_OK;
			}
E
Eric Dumazet 已提交
5031
			len = skb_headlen(skb);
5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043
		}
	}

	/* reserve a descriptor for the offload context */
	if ((mss) || (skb->ip_summed == CHECKSUM_PARTIAL))
		count++;
	count++;

	count += TXD_USE_COUNT(len, max_txd_pwr);

	nr_frags = skb_shinfo(skb)->nr_frags;
	for (f = 0; f < nr_frags; f++)
E
Eric Dumazet 已提交
5044
		count += TXD_USE_COUNT(skb_frag_size(&skb_shinfo(skb)->frags[f]),
5045 5046 5047 5048 5049
				       max_txd_pwr);

	if (adapter->hw.mac.tx_pkt_filtering)
		e1000_transfer_dhcp_info(adapter, skb);

5050 5051 5052 5053
	/*
	 * need: count + 2 desc gap to keep tail from touching
	 * head, otherwise try next time
	 */
5054
	if (e1000_maybe_stop_tx(tx_ring, count + 2))
5055 5056
		return NETDEV_TX_BUSY;

5057
	if (vlan_tx_tag_present(skb)) {
5058 5059 5060 5061 5062 5063
		tx_flags |= E1000_TX_FLAGS_VLAN;
		tx_flags |= (vlan_tx_tag_get(skb) << E1000_TX_FLAGS_VLAN_SHIFT);
	}

	first = tx_ring->next_to_use;

5064
	tso = e1000_tso(tx_ring, skb);
5065 5066 5067 5068 5069 5070 5071
	if (tso < 0) {
		dev_kfree_skb_any(skb);
		return NETDEV_TX_OK;
	}

	if (tso)
		tx_flags |= E1000_TX_FLAGS_TSO;
5072
	else if (e1000_tx_csum(tx_ring, skb))
5073 5074
		tx_flags |= E1000_TX_FLAGS_CSUM;

5075 5076
	/*
	 * Old method was to assume IPv4 packet by default if TSO was enabled.
5077
	 * 82571 hardware supports TSO capabilities for IPv6 as well...
5078 5079
	 * no longer assume, we must.
	 */
5080 5081 5082
	if (skb->protocol == htons(ETH_P_IP))
		tx_flags |= E1000_TX_FLAGS_IPV4;

L
Lucas De Marchi 已提交
5083
	/* if count is 0 then mapping error has occurred */
5084
	count = e1000_tx_map(tx_ring, skb, first, max_per_txd, nr_frags, mss);
5085
	if (count) {
5086
		netdev_sent_queue(netdev, skb->len);
5087
		e1000_tx_queue(tx_ring, tx_flags, count);
5088
		/* Make sure there is space in the ring for the next send. */
5089
		e1000_maybe_stop_tx(tx_ring, MAX_SKB_FRAGS + 2);
5090 5091

	} else {
5092
		dev_kfree_skb_any(skb);
5093 5094
		tx_ring->buffer_info[first].time_stamp = 0;
		tx_ring->next_to_use = first;
5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117
	}

	return NETDEV_TX_OK;
}

/**
 * e1000_tx_timeout - Respond to a Tx Hang
 * @netdev: network interface device structure
 **/
static void e1000_tx_timeout(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);

	/* Do the reset outside of interrupt context */
	adapter->tx_timeout_count++;
	schedule_work(&adapter->reset_task);
}

static void e1000_reset_task(struct work_struct *work)
{
	struct e1000_adapter *adapter;
	adapter = container_of(work, struct e1000_adapter, reset_task);

5118 5119 5120 5121
	/* don't run the task if already down */
	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

5122 5123 5124 5125 5126
	if (!((adapter->flags & FLAG_RX_NEEDS_RESTART) &&
	      (adapter->flags & FLAG_RX_RESTART_NOW))) {
		e1000e_dump(adapter);
		e_err("Reset adapter\n");
	}
5127 5128 5129 5130
	e1000e_reinit_locked(adapter);
}

/**
J
Jeff Kirsher 已提交
5131
 * e1000_get_stats64 - Get System Network Statistics
5132
 * @netdev: network interface device structure
J
Jeff Kirsher 已提交
5133
 * @stats: rtnl_link_stats64 pointer
5134 5135 5136
 *
 * Returns the address of the device statistics structure.
 **/
J
Jeff Kirsher 已提交
5137 5138
struct rtnl_link_stats64 *e1000e_get_stats64(struct net_device *netdev,
                                             struct rtnl_link_stats64 *stats)
5139
{
J
Jeff Kirsher 已提交
5140 5141 5142 5143 5144 5145 5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179
	struct e1000_adapter *adapter = netdev_priv(netdev);

	memset(stats, 0, sizeof(struct rtnl_link_stats64));
	spin_lock(&adapter->stats64_lock);
	e1000e_update_stats(adapter);
	/* Fill out the OS statistics structure */
	stats->rx_bytes = adapter->stats.gorc;
	stats->rx_packets = adapter->stats.gprc;
	stats->tx_bytes = adapter->stats.gotc;
	stats->tx_packets = adapter->stats.gptc;
	stats->multicast = adapter->stats.mprc;
	stats->collisions = adapter->stats.colc;

	/* Rx Errors */

	/*
	 * RLEC on some newer hardware can be incorrect so build
	 * our own version based on RUC and ROC
	 */
	stats->rx_errors = adapter->stats.rxerrc +
		adapter->stats.crcerrs + adapter->stats.algnerrc +
		adapter->stats.ruc + adapter->stats.roc +
		adapter->stats.cexterr;
	stats->rx_length_errors = adapter->stats.ruc +
					      adapter->stats.roc;
	stats->rx_crc_errors = adapter->stats.crcerrs;
	stats->rx_frame_errors = adapter->stats.algnerrc;
	stats->rx_missed_errors = adapter->stats.mpc;

	/* Tx Errors */
	stats->tx_errors = adapter->stats.ecol +
				       adapter->stats.latecol;
	stats->tx_aborted_errors = adapter->stats.ecol;
	stats->tx_window_errors = adapter->stats.latecol;
	stats->tx_carrier_errors = adapter->stats.tncrs;

	/* Tx Dropped needs to be maintained elsewhere */

	spin_unlock(&adapter->stats64_lock);
	return stats;
5180 5181 5182 5183 5184 5185 5186 5187 5188 5189 5190 5191 5192 5193
}

/**
 * e1000_change_mtu - Change the Maximum Transfer Unit
 * @netdev: network interface device structure
 * @new_mtu: new value for maximum frame size
 *
 * Returns 0 on success, negative on failure
 **/
static int e1000_change_mtu(struct net_device *netdev, int new_mtu)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;

5194
	/* Jumbo frame support */
5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210
	if (max_frame > ETH_FRAME_LEN + ETH_FCS_LEN) {
		if (!(adapter->flags & FLAG_HAS_JUMBO_FRAMES)) {
			e_err("Jumbo Frames not supported.\n");
			return -EINVAL;
		}

		/*
		 * IP payload checksum (enabled with jumbos/packet-split when
		 * Rx checksum is enabled) and generation of RSS hash is
		 * mutually exclusive in the hardware.
		 */
		if ((netdev->features & NETIF_F_RXCSUM) &&
		    (netdev->features & NETIF_F_RXHASH)) {
			e_err("Jumbo frames cannot be enabled when both receive checksum offload and receive hashing are enabled.  Disable one of the receive offload features before enabling jumbos.\n");
			return -EINVAL;
		}
5211 5212
	}

5213 5214 5215 5216
	/* Supported frame sizes */
	if ((new_mtu < ETH_ZLEN + ETH_FCS_LEN + VLAN_HLEN) ||
	    (max_frame > adapter->max_hw_frame_size)) {
		e_err("Unsupported MTU setting\n");
5217 5218 5219
		return -EINVAL;
	}

5220 5221 5222 5223
	/* Jumbo frame workaround on 82579 requires CRC be stripped */
	if ((adapter->hw.mac.type == e1000_pch2lan) &&
	    !(adapter->flags2 & FLAG2_CRC_STRIPPING) &&
	    (new_mtu > ETH_DATA_LEN)) {
5224
		e_err("Jumbo Frames not supported on 82579 when CRC stripping is disabled.\n");
5225 5226 5227
		return -EINVAL;
	}

5228 5229 5230 5231 5232 5233 5234 5235
	/* 82573 Errata 17 */
	if (((adapter->hw.mac.type == e1000_82573) ||
	     (adapter->hw.mac.type == e1000_82574)) &&
	    (max_frame > ETH_FRAME_LEN + ETH_FCS_LEN)) {
		adapter->flags2 |= FLAG2_DISABLE_ASPM_L1;
		e1000e_disable_aspm(adapter->pdev, PCIE_LINK_STATE_L1);
	}

5236
	while (test_and_set_bit(__E1000_RESETTING, &adapter->state))
5237
		usleep_range(1000, 2000);
5238
	/* e1000e_down -> e1000e_reset dependent on max_frame_size & mtu */
5239
	adapter->max_frame_size = max_frame;
5240 5241
	e_info("changing MTU from %d to %d\n", netdev->mtu, new_mtu);
	netdev->mtu = new_mtu;
5242 5243 5244
	if (netif_running(netdev))
		e1000e_down(adapter);

5245 5246
	/*
	 * NOTE: netdev_alloc_skb reserves 16 bytes, and typically NET_IP_ALIGN
5247 5248
	 * means we reserve 2 more, this pushes us to allocate from the next
	 * larger slab size.
5249
	 * i.e. RXBUFFER_2048 --> size-4096 slab
5250 5251
	 * However with the new *_jumbo_rx* routines, jumbo receives will use
	 * fragmented skbs
5252
	 */
5253

5254
	if (max_frame <= 2048)
5255 5256 5257 5258 5259 5260 5261 5262
		adapter->rx_buffer_len = 2048;
	else
		adapter->rx_buffer_len = 4096;

	/* adjust allocation if LPE protects us, and we aren't using SBP */
	if ((max_frame == ETH_FRAME_LEN + ETH_FCS_LEN) ||
	     (max_frame == ETH_FRAME_LEN + VLAN_HLEN + ETH_FCS_LEN))
		adapter->rx_buffer_len = ETH_FRAME_LEN + VLAN_HLEN
5263
					 + ETH_FCS_LEN;
5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280

	if (netif_running(netdev))
		e1000e_up(adapter);
	else
		e1000e_reset(adapter);

	clear_bit(__E1000_RESETTING, &adapter->state);

	return 0;
}

static int e1000_mii_ioctl(struct net_device *netdev, struct ifreq *ifr,
			   int cmd)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct mii_ioctl_data *data = if_mii(ifr);

5281
	if (adapter->hw.phy.media_type != e1000_media_type_copper)
5282 5283 5284 5285 5286 5287 5288
		return -EOPNOTSUPP;

	switch (cmd) {
	case SIOCGMIIPHY:
		data->phy_id = adapter->hw.phy.addr;
		break;
	case SIOCGMIIREG:
5289 5290
		e1000_phy_read_status(adapter);

5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301 5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321 5322
		switch (data->reg_num & 0x1F) {
		case MII_BMCR:
			data->val_out = adapter->phy_regs.bmcr;
			break;
		case MII_BMSR:
			data->val_out = adapter->phy_regs.bmsr;
			break;
		case MII_PHYSID1:
			data->val_out = (adapter->hw.phy.id >> 16);
			break;
		case MII_PHYSID2:
			data->val_out = (adapter->hw.phy.id & 0xFFFF);
			break;
		case MII_ADVERTISE:
			data->val_out = adapter->phy_regs.advertise;
			break;
		case MII_LPA:
			data->val_out = adapter->phy_regs.lpa;
			break;
		case MII_EXPANSION:
			data->val_out = adapter->phy_regs.expansion;
			break;
		case MII_CTRL1000:
			data->val_out = adapter->phy_regs.ctrl1000;
			break;
		case MII_STAT1000:
			data->val_out = adapter->phy_regs.stat1000;
			break;
		case MII_ESTATUS:
			data->val_out = adapter->phy_regs.estatus;
			break;
		default:
5323 5324 5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336 5337 5338 5339 5340 5341 5342 5343 5344
			return -EIO;
		}
		break;
	case SIOCSMIIREG:
	default:
		return -EOPNOTSUPP;
	}
	return 0;
}

static int e1000_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
{
	switch (cmd) {
	case SIOCGMIIPHY:
	case SIOCGMIIREG:
	case SIOCSMIIREG:
		return e1000_mii_ioctl(netdev, ifr, cmd);
	default:
		return -EOPNOTSUPP;
	}
}

5345 5346 5347 5348
static int e1000_init_phy_wakeup(struct e1000_adapter *adapter, u32 wufc)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 i, mac_reg;
5349
	u16 phy_reg, wuc_enable;
5350 5351 5352
	int retval = 0;

	/* copy MAC RARs to PHY RARs */
5353
	e1000_copy_rx_addrs_to_phy_ich8lan(hw);
5354

5355 5356 5357 5358 5359 5360 5361 5362 5363 5364 5365 5366
	retval = hw->phy.ops.acquire(hw);
	if (retval) {
		e_err("Could not acquire PHY\n");
		return retval;
	}

	/* Enable access to wakeup registers on and set page to BM_WUC_PAGE */
	retval = e1000_enable_phy_wakeup_reg_access_bm(hw, &wuc_enable);
	if (retval)
		goto out;

	/* copy MAC MTA to PHY MTA - only needed for pchlan */
5367 5368
	for (i = 0; i < adapter->hw.mac.mta_reg_count; i++) {
		mac_reg = E1000_READ_REG_ARRAY(hw, E1000_MTA, i);
5369 5370 5371 5372
		hw->phy.ops.write_reg_page(hw, BM_MTA(i),
					   (u16)(mac_reg & 0xFFFF));
		hw->phy.ops.write_reg_page(hw, BM_MTA(i) + 1,
					   (u16)((mac_reg >> 16) & 0xFFFF));
5373 5374 5375
	}

	/* configure PHY Rx Control register */
5376
	hw->phy.ops.read_reg_page(&adapter->hw, BM_RCTL, &phy_reg);
5377 5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392
	mac_reg = er32(RCTL);
	if (mac_reg & E1000_RCTL_UPE)
		phy_reg |= BM_RCTL_UPE;
	if (mac_reg & E1000_RCTL_MPE)
		phy_reg |= BM_RCTL_MPE;
	phy_reg &= ~(BM_RCTL_MO_MASK);
	if (mac_reg & E1000_RCTL_MO_3)
		phy_reg |= (((mac_reg & E1000_RCTL_MO_3) >> E1000_RCTL_MO_SHIFT)
				<< BM_RCTL_MO_SHIFT);
	if (mac_reg & E1000_RCTL_BAM)
		phy_reg |= BM_RCTL_BAM;
	if (mac_reg & E1000_RCTL_PMCF)
		phy_reg |= BM_RCTL_PMCF;
	mac_reg = er32(CTRL);
	if (mac_reg & E1000_CTRL_RFCE)
		phy_reg |= BM_RCTL_RFCE;
5393
	hw->phy.ops.write_reg_page(&adapter->hw, BM_RCTL, phy_reg);
5394 5395 5396 5397 5398 5399

	/* enable PHY wakeup in MAC register */
	ew32(WUFC, wufc);
	ew32(WUC, E1000_WUC_PHY_WAKE | E1000_WUC_PME_EN);

	/* configure and enable PHY wakeup in PHY registers */
5400 5401
	hw->phy.ops.write_reg_page(&adapter->hw, BM_WUFC, wufc);
	hw->phy.ops.write_reg_page(&adapter->hw, BM_WUC, E1000_WUC_PME_EN);
5402 5403

	/* activate PHY wakeup */
5404 5405
	wuc_enable |= BM_WUC_ENABLE_BIT | BM_WUC_HOST_WU_BIT;
	retval = e1000_disable_phy_wakeup_reg_access_bm(hw, &wuc_enable);
5406 5407 5408
	if (retval)
		e_err("Could not set PHY Host Wakeup bit\n");
out:
5409
	hw->phy.ops.release(hw);
5410 5411 5412 5413

	return retval;
}

5414 5415
static int __e1000_shutdown(struct pci_dev *pdev, bool *enable_wake,
			    bool runtime)
5416 5417 5418 5419 5420
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl, ctrl_ext, rctl, status;
5421 5422
	/* Runtime suspend should only enable wakeup for link changes */
	u32 wufc = runtime ? E1000_WUFC_LNKC : adapter->wol;
5423 5424 5425 5426 5427 5428 5429 5430 5431
	int retval = 0;

	netif_device_detach(netdev);

	if (netif_running(netdev)) {
		WARN_ON(test_bit(__E1000_RESETTING, &adapter->state));
		e1000e_down(adapter);
		e1000_free_irq(adapter);
	}
5432
	e1000e_reset_interrupt_capability(adapter);
5433 5434 5435 5436 5437 5438 5439 5440 5441 5442 5443

	retval = pci_save_state(pdev);
	if (retval)
		return retval;

	status = er32(STATUS);
	if (status & E1000_STATUS_LU)
		wufc &= ~E1000_WUFC_LNKC;

	if (wufc) {
		e1000_setup_rctl(adapter);
5444
		e1000e_set_rx_mode(netdev);
5445 5446 5447 5448 5449 5450 5451 5452 5453 5454 5455 5456 5457

		/* turn on all-multi mode if wake on multicast is enabled */
		if (wufc & E1000_WUFC_MC) {
			rctl = er32(RCTL);
			rctl |= E1000_RCTL_MPE;
			ew32(RCTL, rctl);
		}

		ctrl = er32(CTRL);
		/* advertise wake from D3Cold */
		#define E1000_CTRL_ADVD3WUC 0x00100000
		/* phy power management enable */
		#define E1000_CTRL_EN_PHY_PWR_MGMT 0x00200000
5458 5459 5460
		ctrl |= E1000_CTRL_ADVD3WUC;
		if (!(adapter->flags2 & FLAG2_HAS_PHY_WAKEUP))
			ctrl |= E1000_CTRL_EN_PHY_PWR_MGMT;
5461 5462
		ew32(CTRL, ctrl);

5463 5464 5465
		if (adapter->hw.phy.media_type == e1000_media_type_fiber ||
		    adapter->hw.phy.media_type ==
		    e1000_media_type_internal_serdes) {
5466 5467
			/* keep the laser running in D3 */
			ctrl_ext = er32(CTRL_EXT);
5468
			ctrl_ext |= E1000_CTRL_EXT_SDP3_DATA;
5469 5470 5471
			ew32(CTRL_EXT, ctrl_ext);
		}

5472
		if (adapter->flags & FLAG_IS_ICH)
5473
			e1000_suspend_workarounds_ich8lan(&adapter->hw);
5474

5475 5476 5477
		/* Allow time for pending master requests to run */
		e1000e_disable_pcie_master(&adapter->hw);

5478
		if (adapter->flags2 & FLAG2_HAS_PHY_WAKEUP) {
5479 5480 5481 5482 5483 5484 5485 5486 5487
			/* enable wakeup by the PHY */
			retval = e1000_init_phy_wakeup(adapter, wufc);
			if (retval)
				return retval;
		} else {
			/* enable wakeup by the MAC */
			ew32(WUFC, wufc);
			ew32(WUC, E1000_WUC_PME_EN);
		}
5488 5489 5490 5491 5492
	} else {
		ew32(WUC, 0);
		ew32(WUFC, 0);
	}

5493 5494
	*enable_wake = !!wufc;

5495
	/* make sure adapter isn't asleep if manageability is enabled */
5496 5497
	if ((adapter->flags & FLAG_MNG_PT_ENABLED) ||
	    (hw->mac.ops.check_mng_mode(hw)))
5498
		*enable_wake = true;
5499 5500 5501 5502

	if (adapter->hw.phy.type == e1000_phy_igp_3)
		e1000e_igp3_phy_powerdown_workaround_ich8lan(&adapter->hw);

5503 5504 5505 5506
	/*
	 * Release control of h/w to f/w.  If f/w is AMT enabled, this
	 * would have already happened in close and is redundant.
	 */
5507
	e1000e_release_hw_control(adapter);
5508 5509 5510

	pci_disable_device(pdev);

5511 5512 5513 5514 5515 5516 5517 5518 5519 5520 5521 5522 5523 5524 5525 5526 5527 5528 5529 5530
	return 0;
}

static void e1000_power_off(struct pci_dev *pdev, bool sleep, bool wake)
{
	if (sleep && wake) {
		pci_prepare_to_sleep(pdev);
		return;
	}

	pci_wake_from_d3(pdev, wake);
	pci_set_power_state(pdev, PCI_D3hot);
}

static void e1000_complete_shutdown(struct pci_dev *pdev, bool sleep,
                                    bool wake)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

5531 5532 5533 5534 5535 5536 5537 5538
	/*
	 * The pci-e switch on some quad port adapters will report a
	 * correctable error when the MAC transitions from D0 to D3.  To
	 * prevent this we need to mask off the correctable errors on the
	 * downstream port of the pci-e switch.
	 */
	if (adapter->flags & FLAG_IS_QUAD_PORT) {
		struct pci_dev *us_dev = pdev->bus->self;
5539
		int pos = pci_pcie_cap(us_dev);
5540 5541 5542 5543 5544 5545
		u16 devctl;

		pci_read_config_word(us_dev, pos + PCI_EXP_DEVCTL, &devctl);
		pci_write_config_word(us_dev, pos + PCI_EXP_DEVCTL,
		                      (devctl & ~PCI_EXP_DEVCTL_CERE));

5546
		e1000_power_off(pdev, sleep, wake);
5547 5548 5549

		pci_write_config_word(us_dev, pos + PCI_EXP_DEVCTL, devctl);
	} else {
5550
		e1000_power_off(pdev, sleep, wake);
5551
	}
5552 5553
}

5554 5555 5556
#ifdef CONFIG_PCIEASPM
static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
{
5557
	pci_disable_link_state_locked(pdev, state);
5558 5559 5560
}
#else
static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
5561 5562
{
	int pos;
5563
	u16 reg16;
5564 5565

	/*
5566 5567
	 * Both device and parent should have the same ASPM setting.
	 * Disable ASPM in downstream component first and then upstream.
5568
	 */
5569 5570 5571 5572 5573
	pos = pci_pcie_cap(pdev);
	pci_read_config_word(pdev, pos + PCI_EXP_LNKCTL, &reg16);
	reg16 &= ~state;
	pci_write_config_word(pdev, pos + PCI_EXP_LNKCTL, reg16);

5574 5575 5576
	if (!pdev->bus->self)
		return;

5577 5578 5579 5580 5581 5582
	pos = pci_pcie_cap(pdev->bus->self);
	pci_read_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, &reg16);
	reg16 &= ~state;
	pci_write_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, reg16);
}
#endif
5583
static void e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
5584 5585 5586 5587 5588 5589
{
	dev_info(&pdev->dev, "Disabling ASPM %s %s\n",
		 (state & PCIE_LINK_STATE_L0S) ? "L0s" : "",
		 (state & PCIE_LINK_STATE_L1) ? "L1" : "");

	__e1000e_disable_aspm(pdev, state);
5590 5591
}

R
Rafael J. Wysocki 已提交
5592
#ifdef CONFIG_PM
5593
static bool e1000e_pm_ready(struct e1000_adapter *adapter)
5594
{
5595
	return !!adapter->tx_ring->buffer_info;
5596 5597
}

5598
static int __e1000_resume(struct pci_dev *pdev)
5599 5600 5601 5602
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
5603
	u16 aspm_disable_flag = 0;
5604 5605
	u32 err;

5606 5607 5608 5609 5610 5611 5612
	if (adapter->flags2 & FLAG2_DISABLE_ASPM_L0S)
		aspm_disable_flag = PCIE_LINK_STATE_L0S;
	if (adapter->flags2 & FLAG2_DISABLE_ASPM_L1)
		aspm_disable_flag |= PCIE_LINK_STATE_L1;
	if (aspm_disable_flag)
		e1000e_disable_aspm(pdev, aspm_disable_flag);

5613 5614
	pci_set_power_state(pdev, PCI_D0);
	pci_restore_state(pdev);
5615
	pci_save_state(pdev);
T
Taku Izumi 已提交
5616

5617
	e1000e_set_interrupt_capability(adapter);
5618 5619 5620 5621 5622 5623
	if (netif_running(netdev)) {
		err = e1000_request_irq(adapter);
		if (err)
			return err;
	}

5624 5625 5626
	if (hw->mac.type == e1000_pch2lan)
		e1000_resume_workarounds_pchlan(&adapter->hw);

5627
	e1000e_power_up_phy(adapter);
5628 5629 5630 5631 5632 5633 5634 5635 5636 5637 5638 5639

	/* report the system wakeup cause from S3/S4 */
	if (adapter->flags2 & FLAG2_HAS_PHY_WAKEUP) {
		u16 phy_data;

		e1e_rphy(&adapter->hw, BM_WUS, &phy_data);
		if (phy_data) {
			e_info("PHY Wakeup cause - %s\n",
				phy_data & E1000_WUS_EX ? "Unicast Packet" :
				phy_data & E1000_WUS_MC ? "Multicast Packet" :
				phy_data & E1000_WUS_BC ? "Broadcast Packet" :
				phy_data & E1000_WUS_MAG ? "Magic Packet" :
5640 5641
				phy_data & E1000_WUS_LNKC ?
				"Link Status Change" : "other");
5642 5643 5644 5645 5646 5647 5648 5649 5650 5651 5652 5653 5654 5655 5656 5657
		}
		e1e_wphy(&adapter->hw, BM_WUS, ~0);
	} else {
		u32 wus = er32(WUS);
		if (wus) {
			e_info("MAC Wakeup cause - %s\n",
				wus & E1000_WUS_EX ? "Unicast Packet" :
				wus & E1000_WUS_MC ? "Multicast Packet" :
				wus & E1000_WUS_BC ? "Broadcast Packet" :
				wus & E1000_WUS_MAG ? "Magic Packet" :
				wus & E1000_WUS_LNKC ? "Link Status Change" :
				"other");
		}
		ew32(WUS, ~0);
	}

5658 5659
	e1000e_reset(adapter);

5660
	e1000_init_manageability_pt(adapter);
5661 5662 5663 5664 5665 5666

	if (netif_running(netdev))
		e1000e_up(adapter);

	netif_device_attach(netdev);

5667 5668
	/*
	 * If the controller has AMT, do not set DRV_LOAD until the interface
5669
	 * is up.  For all other cases, let the f/w know that the h/w is now
5670 5671
	 * under the control of the driver.
	 */
J
Jesse Brandeburg 已提交
5672
	if (!(adapter->flags & FLAG_HAS_AMT))
5673
		e1000e_get_hw_control(adapter);
5674 5675 5676

	return 0;
}
5677

5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690 5691
#ifdef CONFIG_PM_SLEEP
static int e1000_suspend(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	int retval;
	bool wake;

	retval = __e1000_shutdown(pdev, &wake, false);
	if (!retval)
		e1000_complete_shutdown(pdev, true, wake);

	return retval;
}

5692 5693 5694 5695 5696 5697 5698 5699 5700 5701 5702
static int e1000_resume(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (e1000e_pm_ready(adapter))
		adapter->idle_check = true;

	return __e1000_resume(pdev);
}
5703 5704 5705 5706 5707 5708 5709 5710 5711 5712 5713 5714 5715 5716 5717 5718 5719 5720 5721 5722 5723 5724 5725 5726 5727 5728 5729 5730 5731 5732 5733 5734 5735 5736 5737
#endif /* CONFIG_PM_SLEEP */

#ifdef CONFIG_PM_RUNTIME
static int e1000_runtime_suspend(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (e1000e_pm_ready(adapter)) {
		bool wake;

		__e1000_shutdown(pdev, &wake, true);
	}

	return 0;
}

static int e1000_idle(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (!e1000e_pm_ready(adapter))
		return 0;

	if (adapter->idle_check) {
		adapter->idle_check = false;
		if (!e1000e_has_link(adapter))
			pm_schedule_suspend(dev, MSEC_PER_SEC);
	}

	return -EBUSY;
}
5738 5739 5740 5741 5742 5743 5744 5745 5746 5747 5748 5749 5750

static int e1000_runtime_resume(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (!e1000e_pm_ready(adapter))
		return 0;

	adapter->idle_check = !dev->power.runtime_auto;
	return __e1000_resume(pdev);
}
5751
#endif /* CONFIG_PM_RUNTIME */
R
Rafael J. Wysocki 已提交
5752
#endif /* CONFIG_PM */
5753 5754 5755

static void e1000_shutdown(struct pci_dev *pdev)
{
5756 5757
	bool wake = false;

5758
	__e1000_shutdown(pdev, &wake, false);
5759 5760 5761

	if (system_state == SYSTEM_POWER_OFF)
		e1000_complete_shutdown(pdev, false, wake);
5762 5763 5764
}

#ifdef CONFIG_NET_POLL_CONTROLLER
5765 5766 5767 5768 5769 5770 5771

static irqreturn_t e1000_intr_msix(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (adapter->msix_entries) {
5772 5773
		int vector, msix_irq;

5774 5775 5776 5777 5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791 5792 5793 5794 5795
		vector = 0;
		msix_irq = adapter->msix_entries[vector].vector;
		disable_irq(msix_irq);
		e1000_intr_msix_rx(msix_irq, netdev);
		enable_irq(msix_irq);

		vector++;
		msix_irq = adapter->msix_entries[vector].vector;
		disable_irq(msix_irq);
		e1000_intr_msix_tx(msix_irq, netdev);
		enable_irq(msix_irq);

		vector++;
		msix_irq = adapter->msix_entries[vector].vector;
		disable_irq(msix_irq);
		e1000_msix_other(msix_irq, netdev);
		enable_irq(msix_irq);
	}

	return IRQ_HANDLED;
}

5796 5797 5798 5799 5800 5801 5802 5803 5804
/*
 * Polling 'interrupt' - used by things like netconsole to send skbs
 * without having to re-enable interrupts. It's not called while
 * the interrupt routine is executing.
 */
static void e1000_netpoll(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);

5805 5806 5807 5808 5809 5810 5811 5812 5813 5814 5815 5816 5817 5818 5819
	switch (adapter->int_mode) {
	case E1000E_INT_MODE_MSIX:
		e1000_intr_msix(adapter->pdev->irq, netdev);
		break;
	case E1000E_INT_MODE_MSI:
		disable_irq(adapter->pdev->irq);
		e1000_intr_msi(adapter->pdev->irq, netdev);
		enable_irq(adapter->pdev->irq);
		break;
	default: /* E1000E_INT_MODE_LEGACY */
		disable_irq(adapter->pdev->irq);
		e1000_intr(adapter->pdev->irq, netdev);
		enable_irq(adapter->pdev->irq);
		break;
	}
5820 5821 5822 5823 5824 5825 5826 5827 5828 5829 5830 5831 5832 5833 5834 5835 5836 5837 5838
}
#endif

/**
 * e1000_io_error_detected - called when PCI error is detected
 * @pdev: Pointer to PCI device
 * @state: The current pci connection state
 *
 * This function is called after a PCI bus error affecting
 * this device has been detected.
 */
static pci_ers_result_t e1000_io_error_detected(struct pci_dev *pdev,
						pci_channel_state_t state)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	netif_device_detach(netdev);

5839 5840 5841
	if (state == pci_channel_io_perm_failure)
		return PCI_ERS_RESULT_DISCONNECT;

5842 5843 5844 5845 5846 5847 5848 5849 5850 5851 5852 5853 5854 5855 5856 5857 5858 5859 5860 5861
	if (netif_running(netdev))
		e1000e_down(adapter);
	pci_disable_device(pdev);

	/* Request a slot slot reset. */
	return PCI_ERS_RESULT_NEED_RESET;
}

/**
 * e1000_io_slot_reset - called after the pci bus has been reset.
 * @pdev: Pointer to PCI device
 *
 * Restart the card from scratch, as if from a cold-boot. Implementation
 * resembles the first-half of the e1000_resume routine.
 */
static pci_ers_result_t e1000_io_slot_reset(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
5862
	u16 aspm_disable_flag = 0;
T
Taku Izumi 已提交
5863
	int err;
J
Jesse Brandeburg 已提交
5864
	pci_ers_result_t result;
5865

5866 5867
	if (adapter->flags2 & FLAG2_DISABLE_ASPM_L0S)
		aspm_disable_flag = PCIE_LINK_STATE_L0S;
5868
	if (adapter->flags2 & FLAG2_DISABLE_ASPM_L1)
5869 5870 5871 5872
		aspm_disable_flag |= PCIE_LINK_STATE_L1;
	if (aspm_disable_flag)
		e1000e_disable_aspm(pdev, aspm_disable_flag);

5873
	err = pci_enable_device_mem(pdev);
T
Taku Izumi 已提交
5874
	if (err) {
5875 5876
		dev_err(&pdev->dev,
			"Cannot re-enable PCI device after reset.\n");
J
Jesse Brandeburg 已提交
5877 5878 5879
		result = PCI_ERS_RESULT_DISCONNECT;
	} else {
		pci_set_master(pdev);
5880
		pdev->state_saved = true;
J
Jesse Brandeburg 已提交
5881
		pci_restore_state(pdev);
5882

J
Jesse Brandeburg 已提交
5883 5884
		pci_enable_wake(pdev, PCI_D3hot, 0);
		pci_enable_wake(pdev, PCI_D3cold, 0);
5885

J
Jesse Brandeburg 已提交
5886 5887 5888 5889
		e1000e_reset(adapter);
		ew32(WUS, ~0);
		result = PCI_ERS_RESULT_RECOVERED;
	}
5890

J
Jesse Brandeburg 已提交
5891 5892 5893
	pci_cleanup_aer_uncorrect_error_status(pdev);

	return result;
5894 5895 5896 5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908
}

/**
 * e1000_io_resume - called when traffic can start flowing again.
 * @pdev: Pointer to PCI device
 *
 * This callback is called when the error recovery driver tells us that
 * its OK to resume normal operation. Implementation resembles the
 * second-half of the e1000_resume routine.
 */
static void e1000_io_resume(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

5909
	e1000_init_manageability_pt(adapter);
5910 5911 5912 5913 5914 5915 5916 5917 5918 5919 5920

	if (netif_running(netdev)) {
		if (e1000e_up(adapter)) {
			dev_err(&pdev->dev,
				"can't bring device back up after reset\n");
			return;
		}
	}

	netif_device_attach(netdev);

5921 5922
	/*
	 * If the controller has AMT, do not set DRV_LOAD until the interface
5923
	 * is up.  For all other cases, let the f/w know that the h/w is now
5924 5925
	 * under the control of the driver.
	 */
J
Jesse Brandeburg 已提交
5926
	if (!(adapter->flags & FLAG_HAS_AMT))
5927
		e1000e_get_hw_control(adapter);
5928 5929 5930 5931 5932 5933 5934

}

static void e1000_print_device_info(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct net_device *netdev = adapter->netdev;
5935 5936
	u32 ret_val;
	u8 pba_str[E1000_PBANUM_LENGTH];
5937 5938

	/* print bus type/speed/width info */
5939
	e_info("(PCI Express:2.5GT/s:%s) %pM\n",
5940 5941 5942 5943
	       /* bus width */
	       ((hw->bus.width == e1000_bus_width_pcie_x4) ? "Width x4" :
	        "Width x1"),
	       /* MAC address */
J
Johannes Berg 已提交
5944
	       netdev->dev_addr);
5945 5946
	e_info("Intel(R) PRO/%s Network Connection\n",
	       (hw->phy.type == e1000_phy_ife) ? "10/100" : "1000");
5947 5948 5949
	ret_val = e1000_read_pba_string_generic(hw, pba_str,
						E1000_PBANUM_LENGTH);
	if (ret_val)
5950
		strlcpy((char *)pba_str, "Unknown", sizeof(pba_str));
5951 5952
	e_info("MAC: %d, PHY: %d, PBA No: %s\n",
	       hw->mac.type, hw->phy.type, pba_str);
5953 5954
}

5955 5956 5957 5958 5959 5960 5961 5962 5963 5964
static void e1000_eeprom_checks(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	int ret_val;
	u16 buf = 0;

	if (hw->mac.type != e1000_82573)
		return;

	ret_val = e1000_read_nvm(hw, NVM_INIT_CONTROL2_REG, 1, &buf);
5965
	if (!ret_val && (!(le16_to_cpu(buf) & (1 << 0)))) {
5966
		/* Deep Smart Power Down (DSPD) */
5967 5968
		dev_warn(&adapter->pdev->dev,
			 "Warning: detected DSPD enabled in EEPROM\n");
5969 5970 5971
	}
}

5972
static int e1000_set_features(struct net_device *netdev,
5973
			      netdev_features_t features)
5974 5975
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
5976
	netdev_features_t changed = features ^ netdev->features;
5977 5978 5979 5980 5981

	if (changed & (NETIF_F_TSO | NETIF_F_TSO6))
		adapter->flags |= FLAG_TSO_FORCE;

	if (!(changed & (NETIF_F_HW_VLAN_RX | NETIF_F_HW_VLAN_TX |
5982
			 NETIF_F_RXCSUM | NETIF_F_RXHASH)))
5983 5984
		return 0;

5985 5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997
	/*
	 * IP payload checksum (enabled with jumbos/packet-split when Rx
	 * checksum is enabled) and generation of RSS hash is mutually
	 * exclusive in the hardware.
	 */
	if (adapter->rx_ps_pages &&
	    (features & NETIF_F_RXCSUM) && (features & NETIF_F_RXHASH)) {
		e_err("Enabling both receive checksum offload and receive hashing is not possible with jumbo frames.  Disable jumbos or enable only one of the receive offload features.\n");
		return -EINVAL;
	}

	netdev->features = features;

5998 5999 6000 6001 6002 6003 6004 6005
	if (netif_running(netdev))
		e1000e_reinit_locked(adapter);
	else
		e1000e_reset(adapter);

	return 0;
}

6006 6007 6008
static const struct net_device_ops e1000e_netdev_ops = {
	.ndo_open		= e1000_open,
	.ndo_stop		= e1000_close,
6009
	.ndo_start_xmit		= e1000_xmit_frame,
J
Jeff Kirsher 已提交
6010
	.ndo_get_stats64	= e1000e_get_stats64,
6011
	.ndo_set_rx_mode	= e1000e_set_rx_mode,
6012 6013 6014 6015 6016 6017 6018 6019 6020 6021 6022
	.ndo_set_mac_address	= e1000_set_mac,
	.ndo_change_mtu		= e1000_change_mtu,
	.ndo_do_ioctl		= e1000_ioctl,
	.ndo_tx_timeout		= e1000_tx_timeout,
	.ndo_validate_addr	= eth_validate_addr,

	.ndo_vlan_rx_add_vid	= e1000_vlan_rx_add_vid,
	.ndo_vlan_rx_kill_vid	= e1000_vlan_rx_kill_vid,
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= e1000_netpoll,
#endif
6023
	.ndo_set_features = e1000_set_features,
6024 6025
};

6026 6027 6028 6029 6030 6031 6032 6033 6034 6035 6036 6037 6038 6039 6040 6041 6042 6043
/**
 * e1000_probe - Device Initialization Routine
 * @pdev: PCI device information struct
 * @ent: entry in e1000_pci_tbl
 *
 * Returns 0 on success, negative on failure
 *
 * e1000_probe initializes an adapter identified by a pci_dev structure.
 * The OS initialization, configuring of the adapter private structure,
 * and a hardware reset occur.
 **/
static int __devinit e1000_probe(struct pci_dev *pdev,
				 const struct pci_device_id *ent)
{
	struct net_device *netdev;
	struct e1000_adapter *adapter;
	struct e1000_hw *hw;
	const struct e1000_info *ei = e1000_info_tbl[ent->driver_data];
6044 6045
	resource_size_t mmio_start, mmio_len;
	resource_size_t flash_start, flash_len;
6046 6047

	static int cards_found;
6048
	u16 aspm_disable_flag = 0;
6049 6050 6051 6052
	int i, err, pci_using_dac;
	u16 eeprom_data = 0;
	u16 eeprom_apme_mask = E1000_EEPROM_APME;

6053 6054
	if (ei->flags2 & FLAG2_DISABLE_ASPM_L0S)
		aspm_disable_flag = PCIE_LINK_STATE_L0S;
6055
	if (ei->flags2 & FLAG2_DISABLE_ASPM_L1)
6056 6057 6058
		aspm_disable_flag |= PCIE_LINK_STATE_L1;
	if (aspm_disable_flag)
		e1000e_disable_aspm(pdev, aspm_disable_flag);
T
Taku Izumi 已提交
6059

6060
	err = pci_enable_device_mem(pdev);
6061 6062 6063 6064
	if (err)
		return err;

	pci_using_dac = 0;
6065
	err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
6066
	if (!err) {
6067
		err = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
6068 6069 6070
		if (!err)
			pci_using_dac = 1;
	} else {
6071
		err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
6072
		if (err) {
6073 6074
			err = dma_set_coherent_mask(&pdev->dev,
						    DMA_BIT_MASK(32));
6075
			if (err) {
6076
				dev_err(&pdev->dev, "No usable DMA configuration, aborting\n");
6077 6078 6079 6080 6081
				goto err_dma;
			}
		}
	}

6082
	err = pci_request_selected_regions_exclusive(pdev,
6083 6084
	                                  pci_select_bars(pdev, IORESOURCE_MEM),
	                                  e1000e_driver_name);
6085 6086 6087
	if (err)
		goto err_pci_reg;

6088
	/* AER (Advanced Error Reporting) hooks */
6089
	pci_enable_pcie_error_reporting(pdev);
6090

6091
	pci_set_master(pdev);
6092 6093 6094 6095
	/* PCI config space info */
	err = pci_save_state(pdev);
	if (err)
		goto err_alloc_etherdev;
6096 6097 6098 6099 6100 6101 6102 6103

	err = -ENOMEM;
	netdev = alloc_etherdev(sizeof(struct e1000_adapter));
	if (!netdev)
		goto err_alloc_etherdev;

	SET_NETDEV_DEV(netdev, &pdev->dev);

6104 6105
	netdev->irq = pdev->irq;

6106 6107 6108 6109 6110 6111 6112 6113
	pci_set_drvdata(pdev, netdev);
	adapter = netdev_priv(netdev);
	hw = &adapter->hw;
	adapter->netdev = netdev;
	adapter->pdev = pdev;
	adapter->ei = ei;
	adapter->pba = ei->pba;
	adapter->flags = ei->flags;
J
Jeff Kirsher 已提交
6114
	adapter->flags2 = ei->flags2;
6115 6116
	adapter->hw.adapter = adapter;
	adapter->hw.mac.type = ei->mac;
6117
	adapter->max_hw_frame_size = ei->max_hw_frame_size;
6118 6119 6120 6121 6122 6123 6124 6125 6126 6127 6128 6129 6130 6131 6132 6133 6134 6135 6136 6137
	adapter->msg_enable = (1 << NETIF_MSG_DRV | NETIF_MSG_PROBE) - 1;

	mmio_start = pci_resource_start(pdev, 0);
	mmio_len = pci_resource_len(pdev, 0);

	err = -EIO;
	adapter->hw.hw_addr = ioremap(mmio_start, mmio_len);
	if (!adapter->hw.hw_addr)
		goto err_ioremap;

	if ((adapter->flags & FLAG_HAS_FLASH) &&
	    (pci_resource_flags(pdev, 1) & IORESOURCE_MEM)) {
		flash_start = pci_resource_start(pdev, 1);
		flash_len = pci_resource_len(pdev, 1);
		adapter->hw.flash_address = ioremap(flash_start, flash_len);
		if (!adapter->hw.flash_address)
			goto err_flashmap;
	}

	/* construct the net_device struct */
6138
	netdev->netdev_ops		= &e1000e_netdev_ops;
6139 6140 6141
	e1000e_set_ethtool_ops(netdev);
	netdev->watchdog_timeo		= 5 * HZ;
	netif_napi_add(netdev, &adapter->napi, e1000_clean, 64);
6142
	strlcpy(netdev->name, pci_name(pdev), sizeof(netdev->name));
6143 6144 6145 6146 6147 6148

	netdev->mem_start = mmio_start;
	netdev->mem_end = mmio_start + mmio_len;

	adapter->bd_number = cards_found++;

6149 6150
	e1000e_check_options(adapter);

6151 6152 6153 6154 6155 6156 6157 6158 6159
	/* setup adapter struct */
	err = e1000_sw_init(adapter);
	if (err)
		goto err_sw_init;

	memcpy(&hw->mac.ops, ei->mac_ops, sizeof(hw->mac.ops));
	memcpy(&hw->nvm.ops, ei->nvm_ops, sizeof(hw->nvm.ops));
	memcpy(&hw->phy.ops, ei->phy_ops, sizeof(hw->phy.ops));

J
Jeff Kirsher 已提交
6160
	err = ei->get_variants(adapter);
6161 6162 6163
	if (err)
		goto err_hw_init;

6164 6165 6166 6167
	if ((adapter->flags & FLAG_IS_ICH) &&
	    (adapter->flags & FLAG_READ_ONLY_NVM))
		e1000e_write_protect_nvm_ich8lan(&adapter->hw);

6168 6169
	hw->mac.ops.get_bus_info(&adapter->hw);

6170
	adapter->hw.phy.autoneg_wait_to_complete = 0;
6171 6172

	/* Copper options */
6173
	if (adapter->hw.phy.media_type == e1000_media_type_copper) {
6174 6175 6176 6177 6178 6179
		adapter->hw.phy.mdix = AUTO_ALL_MODES;
		adapter->hw.phy.disable_polarity_correction = 0;
		adapter->hw.phy.ms_type = e1000_ms_hw_default;
	}

	if (e1000_check_reset_block(&adapter->hw))
6180
		e_info("PHY reset is blocked due to SOL/IDER session.\n");
6181

6182 6183 6184 6185 6186 6187
	/* Set initial default active device features */
	netdev->features = (NETIF_F_SG |
			    NETIF_F_HW_VLAN_RX |
			    NETIF_F_HW_VLAN_TX |
			    NETIF_F_TSO |
			    NETIF_F_TSO6 |
6188
			    NETIF_F_RXHASH |
6189 6190 6191 6192 6193
			    NETIF_F_RXCSUM |
			    NETIF_F_HW_CSUM);

	/* Set user-changeable features (subset of all device features) */
	netdev->hw_features = netdev->features;
6194 6195 6196 6197

	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER)
		netdev->features |= NETIF_F_HW_VLAN_FILTER;

6198 6199 6200 6201
	netdev->vlan_features |= (NETIF_F_SG |
				  NETIF_F_TSO |
				  NETIF_F_TSO6 |
				  NETIF_F_HW_CSUM);
6202

6203 6204
	netdev->priv_flags |= IFF_UNICAST_FLT;

6205
	if (pci_using_dac) {
6206
		netdev->features |= NETIF_F_HIGHDMA;
6207 6208
		netdev->vlan_features |= NETIF_F_HIGHDMA;
	}
6209 6210 6211 6212

	if (e1000e_enable_mng_pass_thru(&adapter->hw))
		adapter->flags |= FLAG_MNG_PT_ENABLED;

6213 6214 6215 6216
	/*
	 * before reading the NVM, reset the controller to
	 * put the device in a known good starting state
	 */
6217 6218 6219 6220 6221 6222 6223 6224 6225 6226
	adapter->hw.mac.ops.reset_hw(&adapter->hw);

	/*
	 * systems with ASPM and others may see the checksum fail on the first
	 * attempt. Let's give it a few tries
	 */
	for (i = 0;; i++) {
		if (e1000_validate_nvm_checksum(&adapter->hw) >= 0)
			break;
		if (i == 2) {
6227
			e_err("The NVM Checksum Is Not Valid\n");
6228 6229 6230 6231 6232
			err = -EIO;
			goto err_eeprom;
		}
	}

6233 6234
	e1000_eeprom_checks(adapter);

6235
	/* copy the MAC address */
6236
	if (e1000e_read_mac_addr(&adapter->hw))
6237
		e_err("NVM Read Error while reading MAC address\n");
6238 6239 6240 6241 6242

	memcpy(netdev->dev_addr, adapter->hw.mac.addr, netdev->addr_len);
	memcpy(netdev->perm_addr, adapter->hw.mac.addr, netdev->addr_len);

	if (!is_valid_ether_addr(netdev->perm_addr)) {
J
Johannes Berg 已提交
6243
		e_err("Invalid MAC Address: %pM\n", netdev->perm_addr);
6244 6245 6246 6247 6248
		err = -EIO;
		goto err_eeprom;
	}

	init_timer(&adapter->watchdog_timer);
6249
	adapter->watchdog_timer.function = e1000_watchdog;
6250 6251 6252
	adapter->watchdog_timer.data = (unsigned long) adapter;

	init_timer(&adapter->phy_info_timer);
6253
	adapter->phy_info_timer.function = e1000_update_phy_info;
6254 6255 6256 6257
	adapter->phy_info_timer.data = (unsigned long) adapter;

	INIT_WORK(&adapter->reset_task, e1000_reset_task);
	INIT_WORK(&adapter->watchdog_task, e1000_watchdog_task);
6258 6259
	INIT_WORK(&adapter->downshift_task, e1000e_downshift_workaround);
	INIT_WORK(&adapter->update_phy_task, e1000e_update_phy_task);
6260
	INIT_WORK(&adapter->print_hang_task, e1000_print_hw_hang);
6261 6262 6263

	/* Initialize link parameters. User can change them with ethtool */
	adapter->hw.mac.autoneg = 1;
6264
	adapter->fc_autoneg = true;
6265 6266
	adapter->hw.fc.requested_mode = e1000_fc_default;
	adapter->hw.fc.current_mode = e1000_fc_default;
6267 6268 6269 6270 6271 6272 6273 6274 6275 6276 6277 6278 6279 6280
	adapter->hw.phy.autoneg_advertised = 0x2f;

	/* ring size defaults */
	adapter->rx_ring->count = 256;
	adapter->tx_ring->count = 256;

	/*
	 * Initial Wake on LAN setting - If APM wake is enabled in
	 * the EEPROM, enable the ACPI Magic Packet filter
	 */
	if (adapter->flags & FLAG_APME_IN_WUC) {
		/* APME bit in EEPROM is mapped to WUC.APME */
		eeprom_data = er32(WUC);
		eeprom_apme_mask = E1000_WUC_APME;
6281 6282
		if ((hw->mac.type > e1000_ich10lan) &&
		    (eeprom_data & E1000_WUC_PHY_WAKE))
6283
			adapter->flags2 |= FLAG2_HAS_PHY_WAKEUP;
6284 6285 6286 6287 6288 6289 6290 6291 6292 6293 6294 6295 6296 6297 6298 6299 6300 6301 6302 6303 6304 6305 6306 6307
	} else if (adapter->flags & FLAG_APME_IN_CTRL3) {
		if (adapter->flags & FLAG_APME_CHECK_PORT_B &&
		    (adapter->hw.bus.func == 1))
			e1000_read_nvm(&adapter->hw,
				NVM_INIT_CONTROL3_PORT_B, 1, &eeprom_data);
		else
			e1000_read_nvm(&adapter->hw,
				NVM_INIT_CONTROL3_PORT_A, 1, &eeprom_data);
	}

	/* fetch WoL from EEPROM */
	if (eeprom_data & eeprom_apme_mask)
		adapter->eeprom_wol |= E1000_WUFC_MAG;

	/*
	 * now that we have the eeprom settings, apply the special cases
	 * where the eeprom may be wrong or the board simply won't support
	 * wake on lan on a particular port
	 */
	if (!(adapter->flags & FLAG_HAS_WOL))
		adapter->eeprom_wol = 0;

	/* initialize the wol settings based on the eeprom settings */
	adapter->wol = adapter->eeprom_wol;
6308
	device_set_wakeup_enable(&adapter->pdev->dev, adapter->wol);
6309

6310 6311 6312
	/* save off EEPROM version number */
	e1000_read_nvm(&adapter->hw, 5, 1, &adapter->eeprom_vers);

6313 6314 6315
	/* reset the hardware with the new settings */
	e1000e_reset(adapter);

6316 6317
	/*
	 * If the controller has AMT, do not set DRV_LOAD until the interface
6318
	 * is up.  For all other cases, let the f/w know that the h/w is now
6319 6320
	 * under the control of the driver.
	 */
J
Jesse Brandeburg 已提交
6321
	if (!(adapter->flags & FLAG_HAS_AMT))
6322
		e1000e_get_hw_control(adapter);
6323

6324
	strlcpy(netdev->name, "eth%d", sizeof(netdev->name));
6325 6326 6327 6328
	err = register_netdev(netdev);
	if (err)
		goto err_register;

6329 6330 6331
	/* carrier off reporting is important to ethtool even BEFORE open */
	netif_carrier_off(netdev);

6332 6333
	e1000_print_device_info(adapter);

6334 6335
	if (pci_dev_run_wake(pdev))
		pm_runtime_put_noidle(&pdev->dev);
6336

6337 6338 6339
	return 0;

err_register:
J
Jesse Brandeburg 已提交
6340
	if (!(adapter->flags & FLAG_HAS_AMT))
6341
		e1000e_release_hw_control(adapter);
6342 6343 6344
err_eeprom:
	if (!e1000_check_reset_block(&adapter->hw))
		e1000_phy_hw_reset(&adapter->hw);
J
Jesse Brandeburg 已提交
6345
err_hw_init:
6346 6347 6348
	kfree(adapter->tx_ring);
	kfree(adapter->rx_ring);
err_sw_init:
J
Jesse Brandeburg 已提交
6349 6350
	if (adapter->hw.flash_address)
		iounmap(adapter->hw.flash_address);
6351
	e1000e_reset_interrupt_capability(adapter);
J
Jesse Brandeburg 已提交
6352
err_flashmap:
6353 6354 6355 6356
	iounmap(adapter->hw.hw_addr);
err_ioremap:
	free_netdev(netdev);
err_alloc_etherdev:
6357 6358
	pci_release_selected_regions(pdev,
	                             pci_select_bars(pdev, IORESOURCE_MEM));
6359 6360 6361 6362 6363 6364 6365 6366 6367 6368 6369 6370 6371 6372 6373 6374 6375 6376 6377
err_pci_reg:
err_dma:
	pci_disable_device(pdev);
	return err;
}

/**
 * e1000_remove - Device Removal Routine
 * @pdev: PCI device information struct
 *
 * e1000_remove is called by the PCI subsystem to alert the driver
 * that it should release a PCI device.  The could be caused by a
 * Hot-Plug event, or because the driver is going to be removed from
 * memory.
 **/
static void __devexit e1000_remove(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);
6378 6379
	bool down = test_bit(__E1000_DOWN, &adapter->state);

6380
	/*
6381 6382
	 * The timers may be rescheduled, so explicitly disable them
	 * from being rescheduled.
6383
	 */
6384 6385
	if (!down)
		set_bit(__E1000_DOWN, &adapter->state);
6386 6387 6388
	del_timer_sync(&adapter->watchdog_timer);
	del_timer_sync(&adapter->phy_info_timer);

6389 6390 6391 6392 6393
	cancel_work_sync(&adapter->reset_task);
	cancel_work_sync(&adapter->watchdog_task);
	cancel_work_sync(&adapter->downshift_task);
	cancel_work_sync(&adapter->update_phy_task);
	cancel_work_sync(&adapter->print_hang_task);
6394

6395 6396 6397
	if (!(netdev->flags & IFF_UP))
		e1000_power_down_phy(adapter);

6398 6399 6400
	/* Don't lie to e1000_close() down the road. */
	if (!down)
		clear_bit(__E1000_DOWN, &adapter->state);
6401 6402
	unregister_netdev(netdev);

6403 6404
	if (pci_dev_run_wake(pdev))
		pm_runtime_get_noresume(&pdev->dev);
6405

6406 6407 6408 6409
	/*
	 * Release control of h/w to f/w.  If f/w is AMT enabled, this
	 * would have already happened in close and is redundant.
	 */
6410
	e1000e_release_hw_control(adapter);
6411

6412
	e1000e_reset_interrupt_capability(adapter);
6413 6414 6415 6416 6417 6418
	kfree(adapter->tx_ring);
	kfree(adapter->rx_ring);

	iounmap(adapter->hw.hw_addr);
	if (adapter->hw.flash_address)
		iounmap(adapter->hw.flash_address);
6419 6420
	pci_release_selected_regions(pdev,
	                             pci_select_bars(pdev, IORESOURCE_MEM));
6421 6422 6423

	free_netdev(netdev);

J
Jesse Brandeburg 已提交
6424
	/* AER disable */
6425
	pci_disable_pcie_error_reporting(pdev);
J
Jesse Brandeburg 已提交
6426

6427 6428 6429 6430 6431 6432 6433 6434 6435 6436
	pci_disable_device(pdev);
}

/* PCI Error Recovery (ERS) */
static struct pci_error_handlers e1000_err_handler = {
	.error_detected = e1000_io_error_detected,
	.slot_reset = e1000_io_slot_reset,
	.resume = e1000_io_resume,
};

6437
static DEFINE_PCI_DEVICE_TABLE(e1000_pci_tbl) = {
6438 6439 6440 6441 6442 6443
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_COPPER), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_FIBER), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_QUAD_COPPER), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_QUAD_COPPER_LP), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_QUAD_FIBER), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_SERDES), board_82571 },
6444 6445 6446
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_SERDES_DUAL), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_SERDES_QUAD), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571PT_QUAD_COPPER), board_82571 },
6447

6448 6449 6450 6451
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82572EI), board_82572 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82572EI_COPPER), board_82572 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82572EI_FIBER), board_82572 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82572EI_SERDES), board_82572 },
6452

6453 6454 6455
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82573E), board_82573 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82573E_IAMT), board_82573 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82573L), board_82573 },
6456

6457
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82574L), board_82574 },
6458
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82574LA), board_82574 },
6459
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82583V), board_82583 },
6460

6461 6462 6463 6464 6465 6466 6467 6468
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_80003ES2LAN_COPPER_DPT),
	  board_80003es2lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_80003ES2LAN_COPPER_SPT),
	  board_80003es2lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_80003ES2LAN_SERDES_DPT),
	  board_80003es2lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_80003ES2LAN_SERDES_SPT),
	  board_80003es2lan },
6469

6470 6471 6472 6473 6474 6475 6476
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IFE), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IFE_G), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IFE_GT), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IGP_AMT), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IGP_C), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IGP_M), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IGP_M_AMT), board_ich8lan },
B
Bruce Allan 已提交
6477
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_82567V_3), board_ich8lan },
6478

6479 6480 6481 6482 6483
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IFE), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IFE_G), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IFE_GT), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_AMT), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_C), board_ich9lan },
6484
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_BM), board_ich9lan },
6485 6486 6487 6488 6489 6490 6491
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_M), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_M_AMT), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_M_V), board_ich9lan },

	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_R_BM_LM), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_R_BM_LF), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_R_BM_V), board_ich9lan },
6492

6493 6494
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_D_BM_LM), board_ich10lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_D_BM_LF), board_ich10lan },
6495
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_D_BM_V), board_ich10lan },
6496

6497 6498 6499 6500 6501
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH_M_HV_LM), board_pchlan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH_M_HV_LC), board_pchlan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH_D_HV_DM), board_pchlan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH_D_HV_DC), board_pchlan },

6502 6503 6504
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH2_LV_LM), board_pch2lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH2_LV_V), board_pch2lan },

6505 6506 6507 6508
	{ }	/* terminate list */
};
MODULE_DEVICE_TABLE(pci, e1000_pci_tbl);

R
Rafael J. Wysocki 已提交
6509
#ifdef CONFIG_PM
6510
static const struct dev_pm_ops e1000_pm_ops = {
6511 6512 6513
	SET_SYSTEM_SLEEP_PM_OPS(e1000_suspend, e1000_resume)
	SET_RUNTIME_PM_OPS(e1000_runtime_suspend,
				e1000_runtime_resume, e1000_idle)
6514
};
6515
#endif
6516

6517 6518 6519 6520 6521 6522
/* PCI Device API Driver */
static struct pci_driver e1000_driver = {
	.name     = e1000e_driver_name,
	.id_table = e1000_pci_tbl,
	.probe    = e1000_probe,
	.remove   = __devexit_p(e1000_remove),
R
Rafael J. Wysocki 已提交
6523
#ifdef CONFIG_PM
6524
	.driver.pm = &e1000_pm_ops,
6525 6526 6527 6528 6529 6530 6531 6532 6533 6534 6535 6536 6537 6538
#endif
	.shutdown = e1000_shutdown,
	.err_handler = &e1000_err_handler
};

/**
 * e1000_init_module - Driver Registration Routine
 *
 * e1000_init_module is the first routine called when the driver is
 * loaded. All it does is register with the PCI subsystem.
 **/
static int __init e1000_init_module(void)
{
	int ret;
6539 6540
	pr_info("Intel(R) PRO/1000 Network Driver - %s\n",
		e1000e_driver_version);
B
Bruce Allan 已提交
6541
	pr_info("Copyright(c) 1999 - 2011 Intel Corporation.\n");
6542
	ret = pci_register_driver(&e1000_driver);
6543

6544 6545 6546 6547 6548 6549 6550 6551 6552 6553 6554 6555 6556 6557 6558 6559 6560 6561 6562 6563 6564 6565 6566
	return ret;
}
module_init(e1000_init_module);

/**
 * e1000_exit_module - Driver Exit Cleanup Routine
 *
 * e1000_exit_module is called just before the driver is removed
 * from memory.
 **/
static void __exit e1000_exit_module(void)
{
	pci_unregister_driver(&e1000_driver);
}
module_exit(e1000_exit_module);


MODULE_AUTHOR("Intel Corporation, <linux.nics@intel.com>");
MODULE_DESCRIPTION("Intel(R) PRO/1000 Network Driver");
MODULE_LICENSE("GPL");
MODULE_VERSION(DRV_VERSION);

/* e1000_main.c */